ArticlePDF Available

Ecohydrologic role of solar radiation on landscape evolution

Abstract and Figures

Solar radiation has a clear signature on the spatial organization of ecohydrologic fluxes, vegetation patterns and dynamics, and landscape morphology in semiarid ecosystems. Existing landscape evolution models (LEMs) do not explicitly consider spatially-explicit solar radiation as model forcing. Here, we improve an existing LEM to represent coupled processes of energy, water, and sediment balance for semiarid fluvial catchments. To ground model predictions a study site is selected in central New Mexico where hillslope aspect has a marked influence on vegetation patterns and landscape morphology. Model predictions are corroborated using limited field observations in central NM and other locations with similar conditions. We design a set of comparative LEM simulations to investigate the role of spatially-explicit solar radiation on landscape ecohydro-geomorphic development under different uplift scenarios. Aspect- and network-controls were identified as the two main drivers of soil moisture and vegetation organization on the landscape. Landscape-scale and long-term implications of these short-term ecohdrologic patterns emerged in modeled landscapes. As north facing slopes (NFS) get steeper by continuing uplift they support erosion-resistant denser vegetation cover which leads to further slope steepening until erosion and uplift attains a dynamic equilibrium. Conversely, on north facing slopes (SFS), as slopes grow with uplift, increased solar radiation exposure with slope supports sparser biomass and shallower slopes. At the landscape scale, these differential erosion processes lead to asymmetric development of catchment forms, consistent with regional observations. Understanding of ecohydro-geomorphic evolution will improve to assess the impacts of past and future climates on landscape response and morphology. This article is protected by copyright. All rights reserved.
Content may be subject to copyright.
RESEARCH ARTICLE
10.1002/2014WR016169
Ecohydrologic role of solar radiation on landscape evolution
Omer Yetemen
1,2
, Erkan Istanbulluoglu
1
, J. Homero Flores-Cervantes
1
, Enrique R. Vivoni
3,4
,
and Rafael L. Bras
5
1
Department of Civil and Environmental Engineering, University of Washington, Seattle, Washington, USA,
2
Global
Institute for Water Security, University of Saskatchewan, Saskatoon, Saskatchewan, Canada,
3
School of Earth and Space
Exploration, Arizona State University, Tempe, Arizona, USA,
4
School of Sustainable Engineering and the Built Environment,
Arizona State University, Tempe, Arizona, USA,
5
School of Civil and Environmental Engineering, Georgia Institute of
Technology, Atlanta, Georgia, USA
Abstract Solar radiation has a clear signature on the spatial organization of ecohydrologic fluxes, vegeta-
tion patterns and dynamics, and landscape morphology in semiarid ecosystems. Existing landscape evolu-
tion models (LEMs) do not explicitly consider spatially explicit solar radiation as model forcing. Here, we
improve an existing LEM to represent coupled processes of energy, water, and sediment balance for semi-
arid fluvial catchments. To ground model predictions, a study site is selected in central New Mexico where
hillslope aspect has a marked influence on vegetation patterns and landscape morphology. Model predic-
tions are corroborated using limited field observations in central NM and other locations with similar condi-
tions. We design a set of comparative LEM simulations to investigate the role of spatially explicit solar
radiation on landscape ecohydro-geomorphic development under different uplift scenarios. Aspect-control
and network-control are identified as the two main drivers of soil moisture and vegetation organization on
the landscape. Landscape-scale and long-term implications of these short-term ecohdrologic patterns
emerged in modeled landscapes. As north facing slopes (NFS) get steeper by continuing uplift they support
erosion-resistant denser vegetation cover which leads to further slope steepening until erosion and uplift
attains a dynamic equilibrium. Conversely, on south facing slopes (SFS), as slopes grow with uplift, increased
solar radiation exposure with slope supports sparser biomass and shallower slopes. At the landscape scale,
these differential erosion processes lead to asymmetric development of catchment forms, consistent with
regional observations. Understanding of ecohydrogeomorphic evolution will improve to assess the impacts
of past and future climates on landscape response and morphology.
1. Introduction
Soil and vegetation differences with respect to hillslope orientation have long been recognized. For exam-
ple, the Hebrew Bible (1500 B.C.) describes the ancient Middle Eastern city of Shechem as a hilly country
nestled between the vegetated north-facing slope (NFS) of Mount Gerizim and barren south-facing slopes
(SFS) of Mount Ebal [Hillel, 2006]. In semiarid ecosystems, observed vegetation differences on NFS and SFS
have been largely attributed to the interplay between solar radiation-driven evapotranspiration demand
and rainfall climatology that control local soil moisture, and the spatial organization of plant functional
types and productivity to reduce plant water stress [Ivanov et al., 2008b; Zhou et al., 2013; Flores Cervantes
et al., 2014]. Besides differential vegetation development, increasing evidence shows associations between
slope aspect and soil composition and chemistry [Butler et al., 1986; Kunkel et al., 2011; Ma et al., 2011], as
well as hillslope and landscape geomorphology [Pierce and Colman, 1986; Burnett et al., 2008; Anderson
et al., 2013; West et al., 2014]. In this paper, we examine the role of solar radiation on landscape evolution
through ecohydrologic feedbacks in semiarid ecosystems.
Vegetation is a critical environmental variable that can regulate geomorphic processes [e.g., Dietrich and
Perron, 2006]. Several recent studies demonstrated steeper and less dissected NFS compared to SFS at the
watershed-scale in North America using a range of DEM resolutions [Istanbulluoglu et al., 2008; Poulos et al.,
2012; Guti
errez-Jurado and Vivoni, 2013; West et al., 2014]. Extending comparisons of hillslope morphologies
to the American continents, Poulos et al. [2012] developed a hillslope asymmetry index (HA
NS
) represented
by log10 of the ratio of median slopes () of NFS and SFS. They showed positive HA
NS
between 10N–49N
Key Points:
Energy, water, and biomass
conserved in a landscape evolution
model
Valley asymmetry attributed to
distribution of solar radiation
Uplift enhances role of solar radiation
on catchment development
Correspondence to:
O. Yetemen,
omer.yetemen@usask.ca
Citation:
Yetemen, O., E. Istanbulluoglu,
J. H. Flores-Cervantes, E. R. Vivoni, and
R. L. Bras (2015), Ecohydrologic role of
solar radiation on landscape evolution,
Water Resour. Res.,51, 1127–1157,
doi:10.1002/2014WR016169.
Received 22 JUL 2014
Accepted 13 JAN 2015
Accepted article online 21 JAN 2015
Published online 24 FEB 2015
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1127
Water Resources Research
PUBLICATIONS
latitudes and negative HA
NS
between 15S–40S latitudes, indicating steeper NFS than SFS in the Northern
Hemisphere, and steeper SFS than NFS in the Southern Hemisphere.
Landscape evolution models (LEMs) provide a numerical framework for studying the role of solar radiation
on the evolution of Earth’s surface and its ecological and physical components. Most LEM studies have used
geomorphic transport laws [see Dietrich et al., 2003 for a detailed review] to predict long-term average sedi-
ment flux at the reach-scale, often with the assumption of a year-around steady state runoff rate [e.g., Kirkby
1971, Willgoose et al., 1991; Tucker and Bras, 1998; Perron et al. 2008]. Tucker and Bras [2000] introduced sto-
chastic rainfall forcing in LEM research with distinct pulses of storms, each represented by a rate, duration,
and interstorm period. This advance brought a realistic ‘‘time-clock’’ to catchment evolution determined by
the intermittency of actual storms, which facilitated the development of vegetation life-cycle models within
LEMs. Early studies that coupled vegetation-erosion feedback mechanisms employed simplified vegetation
models, which limited vegetation growth to available space only, such as using a logistic growth curve [e.g.,
Levins, 1969], and related vegetation loss to geomorphic disturbances [Thornes, 1990; Collins et al., 2004;
Istanbulluoglu and Bras, 2005]. These studies predicted strong influences of vegetation dynamics on land-
scape relief, drainage density, and the spatial extent of geomorphic processes in modeled equilibrium land-
scapes. However, absence of both climatic drivers on vegetation dynamics and event-based runoff
generation limits the use of these models to humid landscapes, where water is not a limiting factor for veg-
etation growth.
There are several unique aspects of semiarid ecosystems that need to be represented in LEMs. First of
all, evapotranspiration losses dominate the annual water balance, and both evapotranspiration fluxes
and vegetation dynamics are tightly related to storm characteristics (i.e., intensity, duration, and inter-
storm duration) and hillslope aspect via soil moisture dynamics [Sala and Lauenroth, 1982; Loik et al.,
2004; Porporato et al., 2004; Istanbulluoglu and Bras, 2006; Lauenroth and Bradford, 2009; Dunkerley,
2010; Guti
errez-Jurado et al., 2013]. Second, most semiarid streams are ephemeral, where sediment flux
is driven by high-magnitude, low-frequency, and short-duration flash flood events, generated in tight
relation to spatial soil moisture and vegetation properties [Coppus and Imeson, 2002; Tucker et al., 2006;
Guti
errez-Jurado et al., 2007; Polyakov et al., 2010]. Therefore, a process-based ecohydrogeomorphic LEM
for semiarid ecosystems should realistically represent the spatial and seasonal dynamics of soil moisture,
vegetation biomass, and flood characteristics at the catchment scale. These requirements necessitate
the use of a spatially explicit ecohydrologic modeling approach in a LEM driven by actual storm charac-
teristics, weather conditions, and spatially explicit solar radiation over time scales of catchment geomor-
phic evolution.
Initial mechanistic modeling studies that addressed vegetation-erosion linkages in semiarid climates
coupled formulations of one-dimensional water-balance, moisture-dependent transpiration, and
transpiration-dependent plant growth driven by storm arrivals [Istanbulluoglu and Bras, 2006; Saco et al.,
2007; Collins and Bras, 2008, 2010]. These studies explored some of the well-known empirical concepts and
observations of climate geomorphology, such as the sediment yield-precipitation [Langbein and Schumm,
1958, Wilson, 1973, Summerfield and Hulton, 1994] and drainage density-aridity [Gregory, 1976, Moglen et al.,
1998] relations, and observed banded vegetation patterns [Saco et al., 2007]. With respect to solar forcing,
however, models took a ‘‘flat-earth’’ approach, by assuming spatially uniform solar radiation.
The role of hillslope aspect has been introduced in LEMs in a couple of excellent recent studies. For high-
altitude environments where soil creep is dominated by freeze-thaw processes, Anderson et al. [2013] devel-
oped a 1-D hillslope model by linking a frost-cracking model driven by soil temperature, with models of
regolith production and transport. Soil temperature was parameterized as a function of slope and aspect. At
the Gordon Gulch site of the Boulder Creek CZO, deeper frost development, and resulting more effective
regolith transport on cooler NFS led to southward ridge-top migration and the development of hillslope
asymmetry with shallower and longer NFS and steeper and shorter SFS in the model, consistent with obser-
vations. McGuire et al. [2014] developed a 2-D LEM by coupling models of nonlinear hillslope diffusion and
detachment-limited soil-wash erosion to study topographic asymmetry on some cinder cones in the west-
ern U.S. To capture the influence of observed vegetation biomass that increased in relation to the degree of
‘‘northness’’ on cinder cones, coefficients of hillslope diffusion and soil erodibility were parameterized to
vary linearly with northness. Their modeling experiments underscored the importance of changes in domi-
nant forms of colluvial transport with altitude over Quaternary landscape evolution. However, these studies
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1128
did not explicitly incorporate solar radiation
and energy balance processes in driving of
microclimatic conditions and biophysical
phenomenon.
In this study, we examine the role of solar
radiation on the coupled development of
fluvial landscape morphology and ecohydro-
logic patterns in a semiarid region by cou-
pling water, energy, and mass conservation
laws using the computational framework of
the CHILD (Channel Hillslope Integrated
Landscape Development) LEM [Tucker et al.,
2001a]. We adopt mechanistic models for
solar radiation [e.g., Bras, 1990] and climate-
driven soil moisture and vegetation dynam-
ics [e.g., Montaldo et al., 2005], and couple
ecosystem dynamics with fluvial processes.
By doing so, we take advantage of well-
grounded ecohydrologic model parameter-
izations confirmed by field observations,
and use them to advance LEM theory. In this
paper, we first describe the theory of ecohy-
drogeomorphic landscape evolution, fol-
lowed by the confirmation of the
ecohydrology component of the model in
central New Mexico. We then explore the
role of solar radiation on the coupled ecohydrogeomorphic development of landscapes and ecohydrologic
patterns under varying uplift in a series of numerical simulations developed for central New Mexico condi-
tions. Model findings are corroborated qualitatively with vegetation and landscape observations in central
New Mexico.
2. Model Description
The continuity principles of the land surface energy, water, elevation (sediment mass), and vegetation bio-
mass (currently for a single plant type) are coupled in the CHILD LEM [Tucker et al., 2001a, 2001b] on a spa-
tial model domain represented by Voronoi polygons constructed from a TIN (Triangulated Irregular
Network) network (Figure 1). The coupled system of continuity equations are illustrated here in ‘‘generic’’
forms, the details of which are discussed throughout this section:
Energy :dSh
dt 5RNS;Asp;tðÞ2G2H2kET RN;s;LAIðÞ (1)
Water :nDr
@s
@t5p2rqw2ET RN;s;LAIðÞ (2)
Biomass :dB
dt 5NPP ETðÞ/2kBB(3)
Elevation :@Z
@t5U2rqsd2rqsf sfQo;Vt
ðÞ½ (4)
Energy balance of the land surface can be represented by the rate of change in the surface heat storage
dS
h
/dt (Wm
22
), driven by R
N
: net radiation (Wm
22
), and partitioned into G: ground heat flux (Wm
22
), H: sen-
sible heat flux (Wm
22
), and kET: latent heat flux (Wm
22
) components. Gand Hfluxes are governed by tem-
perature gradients between the surface and the deeper ground, for G, and between the surface and the air
above, for H.kET is the latent energy transferred to the atmosphere through ET, where kis the latent heat
of vaporization (28.34 W d m
22
mm
21
). In the equations above, environmental variables that regulate fluxes
Figure 1. Illustration of the modeled energy, water, and sediment fluxes,
and ecohydrologic state variables in a Voronoi cell used in the CHILD land-
scape evolution model.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1129
are given in parenthesis. Sand
Asp are local topographic slope
and aspect, respectively, tis
time, s[-] is the degree of satura-
tion in the root zone (i.e. volu-
metric soil moisture divided by
porosity, 0 <s1), D
r
[L] is the
effective root depth for plant
water uptake, and LAI [-] is vege-
tation leaf area index estimated
from modeled biomass on the
land surface.
The water balance component
(equation (2)) tracks the changes
in the amount of water in the
soil layer [LT
21
], with a maxi-
mum storage of nD
r
[L], where n
[2] is soil porosity. The compo-
nents of the water balance
equation are the rate of precipi-
tation, p[LT
21
]; divergence of
water flux per unit width, !q
w
[LT
21
] (i.e., difference between
the sum of outgoing overland,
Q
o
, and lateral subsurface, Q
l
,
flows from a cell and the sum of
incoming flows from upstream
cells divided by cell area); and losses due to local ET [LT
21
]. Water and energy balance are coupled through
ET and q
w
, as changes in one regulates the amount of water available for the other.
In most process-based vegetation dynamics models, the rate of change in biomass, B(g DM m
22
), dB/dt is
modeled as the balance between net primary productivity (NPP), driven by ET, as the source term, and first-
order biomass decay represented by a coefficient of k
B
[T
21
] as a sink term [e.g., Montaldo et al., 2005]. NPP is
the net flux of carbon from the atmosphere to green plants, and /is an allocation coefficient of NPP to above-
ground biomass. In CHILD, live and dead constituents of Bare represented separately. Vegetation dynamics is
coupled with water balance through ET.Bregulates the amount of ET,asLAI is estimated from B.
Finally, the continuity of sediment follows the Exner equation that gives the rate of change in elevation, Z,asa
function of uplift, U[LT
21
], and the divergence of volumetric sediment flux per unit width by hillslope diffusion,
!q
sd
[LT
21
], and by fluvial transport, !q
sf
[LT
21
] in a particular direction. In the fluvial component, total vege-
tation cover fraction, V
t
, is used to reduce the efficiency of overland flow effective shear stress, s
f
(Pa), employed
in the sediment detachment and transport capacity equations [e.g., Istanbulluoglu and Bras, 2005, 2006].
Figure 2 shows the integration of the ecohydrology (equations (1), (2), and (3)) and geomorphology (equa-
tion (4)) components in the CHILD LEM. The model coevolves fields of Z(x,y) and B(x,y), driven by pulses of
precipitation (P) with a variable model time step set by interstorm duration, t
b
. In the model, both observed
and statistically generated rainfall data can be used. We assume dS
h
/dt50 to simplify the energy balance
component over interstorm time scales (days to weeks in semiarid climates). Making this assumption allows
us to use a potential evapotranspiration (PET) model, such as the Penman-Monteith combination equation,
as the upper limit of ET, and use PET as external forcing variable. In the model, PET is reduced by a scalar
based on soil moisture availability to calculate actual daily ET [e.g., Brutsaert, 1982; Laio et al., 2001; Montaldo
et al., 2005]. Realistic estimation of actual ET is crucial as plant growth is driven by ET in our model. A solar
radiation ratio, R
solar
(ratio of incoming short-wave radiation on a hillslope element to that of a flat surface)
is multiplied by PET estimated for flat surface, to approximate PET on hillslope elements [Dingman, 2002].
Flow directions (FD) based on the D8 algorithm, local slope (S), and aspect (Asp) are derived from Zto char-
acterize the topographic attributes of the modeled domain. Sand Asp are used to calculate R
solar
, and FD is
used for flow routing. Runoff is routed to the neighbor grid cell in the steepest downhill direction.
Figure 2. Flow chart of the coupled ecohydrology and geomorphology components of
the CHILD LEM driven by tectonic uplift, rainfall characteristics, and potential evapotranspi-
ration (PET) demand. Biomass (B) is the state variable updated and used by both
components.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1130
The model processes Pas an instantaneous pulse, increasing the soil moisture in the root zone, and routes
the excess overland, Q
o
[L
3
T
21
], and subsurface flow, Q
l
[L
3
T
21
], cell by cell. At each cell, s
f
is calculated from
Q
o
and V
t
, and used in detachment and transport capacity models. The Zfield is updated after each storm
with the changes in Zas a result of erosion and deposition (DZ). Erosion is assumed to disrupt vegetation
and cause biomass loss (-DB), creating a positive feedback to fluvial detachment as storm continues or for
the subsequent storm event (Figure 2). ET and vegetation growth/decay are modeled during the interstorm
period. The Bfield is updated by the net change in vegetation biomass (DB) before the arrival of the next
storm. Below model formulations are presented.
2.1. Ecohydrologic Dynamics
This study builds on the ecohydrology [Collins and Bras, 2010] and solar radiation [Flores-Cervantes, 2010]
models previously implemented in CHILD. Currently, the model simulates only grass as a plant functional
type. Grassland ecohydrology is introduced with dynamic live and dead grass biomass state variables
adopted from recent models designed for long-term simulation experiments with minimalistic input varia-
bles sufficient to capture seasonal vegetation dynamics similar to equation (3) [Istanbulluoglu et al., 2012;
Zhou et al., 2013]. Both live and dead biomass types are modeled, as the amount of dead biomass is a criti-
cal regulator of the rate of live biomass generation during the growing season.
Hydrology is modeled as a collection of single-layer storage units, interconnected along flow paths
(Figure 1). Surface runoff is routed cell by cell, and local runoff is generated when the total water
input (local rainfall 1run-on discharge) exceeds the rate of actual infiltration. The rate of global run-
off leaving a model element Q
o
[L
3
T
21
] is:
Qo5max hp1Qsum
in
ac
2Iaiac;0
"# (5)
where p[LT
21
] is rainfall rate, p5P/t
r
(Pis storm depth, t
r
is storm duration), Qsum
in is the sum of run-on discharge
[L
3
T
21
] from upstream sources Qsum
in 5X
k
1
Qin,kis number of upstream cells that drain into a downstream cell,
and a
c
[L
2
]isthecellarea.I
a
[LT
21
] is the actual rate of infiltration. I
a
is constrained by three factors: available
water flux, infiltration capacity, and the available pore space in the root zone [Collins and Bras, 2010]:
Ia5min P1Qin
ac
;Ic;nDr12sðÞ
tr
 (6)
where I
c
[LT
21
] is the infiltration capacity and t
r
is storm duration [T]. In equation (6), no runoff is generated
when I
a
equals the first term. Infiltration (saturation) excess runoff is generated when I
a
equals the second
(third) term. Infiltration capacity at a model element, I
c
, is modeled similar to Dunne et al. [1991], as the
weighted average of I
c
of bare soil, I
c,bare
, and a fully vegetated surface, I
c,veg
:
Ic5Ic;bareð12VtÞ1Ic;veg Vt(7)
With the influence of run-on from upstream cells incorporated into I
a
, a more explicit form of equation (2) is
used for water balance similar to Collins and Bras [2010]:
nZr
@s
@t5Ia1qLp;in2qLp;out 2qL;n2ETa(8)
where q
Lp,in
is the incoming, and q
Lp,out
is the outgoing lateral soil-moisture flux in the direction parallel to
the land surface; and q
L,n
is the normal flux (gravity drainage) [LT
21
], illustrated in Figure 1. q
Lp,out
and q
L,n
are nonzero when s>s
fc
(s
fc
is field capacity), and calculated based on the leakage, L, model of Laio et al.
[2001], resolved into lateral and vertical components. Lateral moisture transfer, q
Lp
, in the direction parallel
to the surface following Cabral et al. [1992]:
qLp5La
rtan S=ð11artan SÞ½ (9)
where Sis local slope angle, a
r
is the anisotropy ratio, which is defined as the ratio of the horizontal to verti-
cal saturated hydraulic conductivity. The normal component of the loss is equal to q
Ln
5L-q
Lp
.q
Ln
is insignifi-
cant comparative to q
Lp
in highly anisotropic medium.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1131
Actual evapotranspiration ET
a
in equation (8) is estimated considering the limitation of root-zone soil mois-
ture as:
ETasðÞ5ETmax s2sw
ðÞ=s2sw
ðÞ;sw<ss(10)
where ET
max
[LT
21
] is the maximum rate of evapotranspiration from a composite soil-vegetation surface; s
w
and s* are soil moisture thresholds for plant wilting point and reduced transpiration under plant water
stress [Laio et al., 2001]. Two other conditions for ET
a
are: ET
a
5ET
max
when s>s*, and ET
a
5E
min
when s<s
wp
.
This models gives a linear increase in ET
a
with soil moisture between s
w
and s*, which plateaus at a constant
rate ET
a
5ET
max
for further increase in s.
We assume that the maximum rate of grass transpiration T
max
(this is more generically defined as PET in Fig-
ure 2) does not exceed the transpiration rate of a reference grass (a well-watered, healthy, live-grass at uni-
form height that completely shades the ground) with a reference leaf area index (m
2
leaf area to m
2
ground
area), LAI
Rmax
of 2.88 [Allen et al., 1989, 1998]. ET
max
at a model element is estimated as the sum of transpira-
tion from grass and evaporation from bare soil [Eagleson, 1978; Brolsma and Bierkens, 2007]:
ETmax 5Tmax
LAIl
LAIRmax
1Es12LAIl
LAIRmax

;LAIlLAIRmax (11)
where LAI
l
is modeled live leaf area index, and the LAI
l
/LAI
Rmax
is the cover fraction of live grass. E
s
[LT
21
]is
the maximum rate of evaporation from bare soil, reduced from T
max
by a coefficient k
s
(E
s
5k
s
T
max
) [e.g., Mut-
ziger et al., 2005; Zhou et al., 2013].
T
max
on sloping model elements, TS
max , is calculated by scaling the T
max
estimated for a flat surface, TF
max ,
with a solar radiation ratio, R
solar
.R
solar
is defined as the ratio of clear-sky radiation on sloped surfaces, RS
cs,to
that of a flat surface, RF
cs:
TS
max 5TF
max Rsolar5TF
max RS
cs
RF
cs
(12a)
TF
max 5TF
ref RN;Ta;v;RH;LAIR
ðÞ (12b)
R
cs
is estimated as a function of day of year (DOY), latitude, local slope, and aspect [Bras, 1990; Dingman,
2002] (see equations in Appendix A). When the model is run with observed meteorological data, the
Penman-Monteith (P-M) equation is used for daily TF
max for reference grass [Allen et al., 1998]. Meteorological
variables employed in the P-M equation are net radiation (R
N
), air temperature (T
a
), wind speed (v), relative
humidity (RH), and LAI
Rmax
(equation (12b)). In long-term geomorphic evolution simulations driven by gen-
erated rainfall, we prescribe TF
max through the use of a cosine function [e.g., Small, 2005], fitted to P-M esti-
mated daily TF
max . In both cases, daily estimates of TF
max are averaged for each interstorm period. The P-M
equation, considerations of radiation balance for estimating R
N
, and the cosine function prescribed for TF
max
for long-term simulations are presented in Appendix B.
There are some limitations of our TS
max model. Equation (12a) neglects the contributions of local variations
in meteorological variables, and diffused and reflected radiation fluxes on TF
max . The latter two terms are
usually relatively small in comparison to direct radiation [e.g., Pierce et al., 2005], unless the modeled topog-
raphy has canyon-like morphology. Assuming uniform meteorological variables is deliberately chosen to
minimize model input requirement, consistent with most catchment-scale distributed hydrologic models
[e.g., Wigmosta et al., 1994].
We illustrate aspect control on radiation distribution, by plotting R
solar
as a function of the day of year (DOY)
and local slope for contrasting N (360) and S (180) aspects for 34N latitude and 107W longitude
(Socorro, NM) (Figure 3). Aspect control on incoming solar radiation is clearly more pronounced during the
fall and winter months (300>DOY>90) because of the shallow gradient of sunlight [Zou et al., 2007]. During
this period, slopes steeper than 10have a greater effect on the distribution of solar radiation. NFS receive
less and SFS receive more solar radiation than a flat surface. In the spring and summer months (growing
season in the region), however, contrasting aspects receive comparable amounts of solar radiation to a flat
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1132
surface and to each other. Observations of solar radiation and ET at our field catchment on NFS and SFS
clearly show this seasonal contrast [Guti
errez-Jurado et al., 2013]. Depending on the slope angle, SFS receive
as much as 5 times greater radiation during the non-growing season than NFS. At the plot scale, with com-
parable local slopes, Guti
errez-Jurado et al. [2013] reported greater rates of ET on SFS slopes during the dry
season than NFS using the Bowen Ratio Energy Balance method. We discuss the implications of spatiotem-
poral variability in solar radiation and rainfall seasonality to ecosystem dynamics later in the paper.
The production of the sum of above-ground and below-ground grass biomass (net primary productivity,
NPP) at the ecosystem scale is related linearly with actual ET through the use of water use efficiency (WUE)
[Swenson and Waring, 2006]:
NPP50:75 12lðÞETaWUE qwx(13)
This is a relatively simple model that represents biomass production by photosynthesis. WUE indicates the
amount of carbon gained for each unit of water lost (kg CO
2
/kg H
2
O), lis the ratio of nighttime to daytime CO
2
exchange, q
w
is water density, and xis a conversion factor of CO
2
to dry biomass (kg DM/kg CO
2
). NPP is parti-
tioned between root and aboveground-live biomass using an allocation coefficient that depends on available
space (equation (14a)). Allocation of biomass to live and dead biomass pools and biomass decay are tracked
using ordinary differential equations as commonly used in vegetation dynamics models [e.g., Sitch et al., 2003].
dBl
dt 5NPP /a2kslBl2ksf nsl Bl(14a)
dBd
dt 5kslBl2kdd nsd Bd(14b)
where B
l
is the aboveground live biomass; and B
d
is the aboveground dead biomass; /
a
is an allocation
coefficient for aboveground live biomass (0</
a
<1); k
sl
and k
dd
are the coefficients for green biomass senes-
cence and dead biomass decay, respectively; k
sf
is coefficient for the drought-induced foliage loss driven by
water stress, n
sl
(0<n
sl
<1); n
sd
is a coefficient for climate-influence on the rate of dead biomass loss [Istan-
bulluoglu et al., 2012]. k
sl
is doubled during the dormancy season to represent unfavorable climatic condi-
tions. Table 1 gives the equations used for /
a
,n
sl
,n
sd
, and the conversion of live and dead LAI (LAI
l
and LAI
d
)
from biomass and the total vegetation cover fraction, V
t
, from total LAI
t
.
The onset and the offset of the growing season are often triggered when a set of environmental conditions
(e.g., soil and air temperature, soil moisture) are satisfied for certain period of time [Cayrol et al., 2000; Sitch
et al., 2003; Ivanov et al., 2008a]. For simplicity, we use the 30 day-averaged T
max
,T
max230
, as a surrogate
variable for climatic favorability [Istanbulluoglu et al., 2012]. The growing season starts when T
max230
is
greater than a growth threshold, GT, and ends when T
max230
gets smaller than a dormancy threshold, DT.
Figure 3. R
solar
plotted for (a) North- and (b) South-facing slopes as a function of day of year (DOY) and local slope. Drying 1 and 2 periods shown in the figures correspond to the two dif-
ferent phases of soil moisture loss, and wetting NAM (North American Monsoon) shows the time and duration of the wet season. These periods were discussed in Figure 13.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1133
2.2. Geomorphic Dynamics
The rate of change in elevation due to flu-
vial processes is set to the lesser of detach-
ment capacity, D
c
[L T
21
], and the
divergence of sediment flux per unit width
!q
sf
[Tucker et al., 2001a]:
@Z
@t52min Dc;
qs;out2Xqs;in

w
ac
2
43
5(15)
where wis width, Rq
s,in
is the total sedi-
ment influx to a model element, and q
s,out
is the outgoing sediment flux from a model
cell assumed at sediment transport
capacity, q
sf
(the magnitude of the vector
q
sf
)[L
2
T
21
]. D
c
gives the maximum rate of
local erosion. We use excess shear stress-
driven parameterization for each flux component described in Table 2 [Istanbulluoglu et al., 2003; Tucker
and Hancock, 2010].
The shear stress that directly acts on soil grains (effective shear stress, s
eff,
(Pa)) [e.g., Einstein and Barbarossa,
1952; Foster, 1982] is used in the excess-shear stress detachment and transport models. s
eff
(Pa) is calculated
from the boundary shear stress scaled by a shear stress partitioning ratio, parameterized using Manning’s
roughness coefficients for bare soil, n
s
, and vegetation n
v
[Laursen, 1958; Istanbulluoglu and Bras, 2005]:
seff 5sb
ns
ns1nv

3=2
(16a)
sb5qwgC0:75 Q0:375
0S0:8125ðns1nvÞ0:375 (16b)
where s
b
is the boundary shear stress (Pa) formulated for a parabolic flow cross section with a shape con-
stant, C50.36, derived from the Manning’s formula for flow velocity.Q
o
is runoff discharge (m
3
s
21
) and Sis
local slope. Dynamic changes in the vegetation cover is introduced through n
v
which is calculate from V
t
in
each iteration of the model (Table 2) [Istanbulluoglu et al., 2003; Istanbulluoglu and Bras, 2005].
In CHILD, vegetation loss due to fluvial scour is represented with a linear loss function driven by excess
shear stress [Collins et al., 2004]:
dVt
dt 52kmVtsf2sc
ðÞ;sf>sc(17)
where k
v
is a vegetation erodibility parameter. It is assumed that loss in live and dead aboveground biomass
is proportional to their initial cover fractions.
2.3. Rainfall Forcing
Model simulations are driven by generated rainfall
using a modified version of the Poisson rectangular
pulse (PRP) rainfall model [Eagleson, 1978]. Each
storm event is represented with constant rainfall
intensity, p, and a storm duration, T
r
, and storms
are separated by an interstorm period, T
b
. A one-
parameter exponential distribution is used to rep-
resent the variability in T
r
and T
b
[Eagleson, 1978].
Dependence between storm depth and its duration
is well-documented [Bonta, 2004; Grayman and
Eagleson, 1969]. To represent this dependence,
storm depths are generated from a two-parameter
Table 1. Equations and References of the Terms in (14a) and (14b)
Ecophysiological Term Equation Source
a
Allocation coefficient
for aboveground
live biomass
/a512LAIg
LAImax 2LAId
 1
Live biomass
water-stress
coefficient nsl 5
0ss
s2s
s2swp

4
swp s
1sswp
s
8
>
>
>
>
<
>
>
>
>
:
2
Climate influence
on dead biomass
loss coefficient
nsd5min Tmax
Tdmax ;1
 1
Live and dead LAI LAIl5cgBl, and LAId5cdBd1
Total LAI LAIt5LAIl1LAId1
Total vegetation
cover fraction
Vt512exp 20:75LAIt
ðÞ 3
a
Sources are as follows: (1) Istanbulluoglu et al. [2012], (2) Laio et al.
[2001]; (3) Lee [1992].
Table 2. Equations of the Terms in Equations (4), (15), (16a),
and (16b)
Geomorphological Term Equation Source
a
Nonlinear hillslope
diffusion
qsd ¼Kd!Z
1ðj!Zj=ScÞ21
Detachment capacity Dc5kdsf2sc
ðÞ
pd 2
Transport capacity qsf 5kfsf2sc
ðÞ
pf 3
Sediment transport
coefficient kf5jffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
gqs=qw21ðÞd3
50
p
qwgqs=qw21ðÞd50
½
pf
4
Vegetation roughness nv5nVR Vt
VR

x5
a
Sources are as follows: (1) Roering et al. [1999]; (2) Nearing
et al. [1999]; (3) Meyer-Peter and M
uller [1948]; (4) Simons and
S¸ent
urk [1992]; and (5) Istanbulluoglu and Bras [2005].
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1134
gamma distribution conditioned by Tr=Tras a dynamic shape parameter that varies with each storm, and
1=P, where Pis the mean storm depth, as the scale parameter of the gamma distribution [Ivanov et al.,
2007]. Rainfall seasonality is simulated by varying the parameters of the distributions based on observed
storm statistics for wet and dry seasons for the region where the model is implemented.
In the PRP model, using the mean rainfall intensity,
p, obtained from rainfall records by dividing the accu-
mulated rainfall during a storm to storm duration often leads to underpredictions of storm runoff, especially
in regions where rainfall is highly variable with high-intensity bursts [Wainwright and Parsons, 2002]. Most
erosive events occur during short-duration high-intensity storm bursts, even though the rainfall event may
continue for longer durations with lower intensities. To obtain both realistic values of storm intensities and
modeled runoff depths, earlier modeling studies have dealt with this limitation using a scaling factor, a,to
calibrate
pand Tras [Collins and Bras, 2010]:
p05a
p;T0r5Tr=a(18)
where p0and T0rare the scaled values of
pand Tr, respectively. To preserve the mean number of storms,
the reduction from storm duration is added to the mean interstorm duration, T0b5Tb1Tr2T0r

.
3. Study Site and Ecohydrologic Model Confirmation
To place model experiments within a relevant ecologic, climatic, and geomorphic context, and corroborate
general patterns in model predictions with observations, a study site is selected at the Sevilleta National
Wildlife Refuge (SNWR) in central New Mexico, where hillslope aspect has a marked influence on vegetation
patterns and landscape morphology (Figure 4). The region receives 250 mm of mean annual precipitation
(MAP). High-intensity and short-duration convective thunderstorms during the North American Monsoon
(NAM, July to September) claim approximately 50% of the MAP [Vivoni et al., 2008], and show high spatial
and interannual variability [Gosz et al., 1995]. In the winter, low-intensity frontal storms with occasional
snow are typical [Milne et al., 2003].
Figure 4 illustrates vegetation and geomorphic characteristics of a catchment selected within the site. The
selected catchment (and others in the region) was incised on coarse alluvial fan deposits of the Plio-
Pleistocene Sierra Ladrones Formation [McMahon, 1998; Connel and McCraw, 2007]. On NFS one-seed Juni-
per (Juniperus monosperma) trees and relatively dense black grama (Bouteloua eriopoda) grass coexist. SFS
host xeric plants consisting of creosote bush (Larrea tridentata) and sparse fluff grass (Erioneuron pulchellum)
[McMahon, 1998; Guti
errez-Jurado et al., 2007]. Soils on NFS contain higher proportions of organic matter,
CaCO
3
, silt, and clay than SFS as a result of higher infiltration rates and deeper infiltration, aeolian sediment
deposition, and root respiration [Guti
errez-Jurado et al., 2006].
Hillslope morphologies of opposing aspects also exhibit pronounced differences. NFS have smooth planar
forms, devoid of any significant fluvial incisions, and are relatively longer than SFS. SFS are highly dissected
with rills and gullies, and exhibit hollow development. Asymmetric development of channel networks has
been observed at this site and in the broader region [Istanbulluoglu et al., 2008; Yetemen et al., 2010].
A plot of local slopes versus contributing drainage areas (S-A) can be used to examine the topographic ram-
ifications of dominant forms of geomorphic transport. Hillslope transport processes such as soil creep and
bioturbation (e.g., equation (1) in Table 2) lead to a positive S-A relationship in which Sincreases with A(i.e.,
a convex hillslope profile), while fluvial processes dominantly acting on the landscape result in concave val-
ley/channel forms leading to a negative S-A relation. Local maxima in a S-A plot between the positive and
negative S-A relations indicates the approximate drainage area required to maintain a fluvial valley network
(i.e., valley head) by providing sufficient discharge for fluvial erosion [e.g., Tucker and Bras, 1998].
The S-A relations for NFS and SFS of the catchment in Figure 4b are obtained from a 10m InSAR DEM (Figure
5), by binning the Sdata for a defined range of Avalues in the basin, and averaging both Sand Adata. Con-
sistent with visual observations in the field, the S-A relation of the basin shows steeper NFS than SFS (5%
steeper over the whole basin). Under a topographic steady state condition, which is likely the case in our
catchment based on Holocene-averaged erosion rate estimates from the
36
Cl cosmogenic dating technique
[Yetemen, 2014], one can make the assumption that erosion on both sides of the valley is driven by an iden-
tical rate of base level fall controlled by the lowering rate of the main channel (i.e., a steady state open-
book catchment model with a channel in the middle of NFS and SFS). A steeper local Sfor a given Awould
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1135
indicate an overall lower efficiency of fluvial activity on NFS as a result of higher biomass, such that S
steepens to maintain a consistent rate of local erosion as the rate of base level fall in the channel. With the
same argument, lower biomass on SFS would lead to enhanced fluvial transport which can effectively erode
hillslopes with shallower slopes than NFS [Istanbulluoglu et al., 2008]. These explanations do not necessarily
require a topographic steady state condition, as in nonsteady conditions and in catchments with threshold
slopes more resistant surfaces to erosion can theoretically grow steeper than more erodible surfaces [e.g.,
Willgoose, 1994; Tucker and Whipple, 2002; Montgomery, 2001; Gasparini et al., 2006]. These interpretations
provide testable hypotheses using the proposed numerical model in this paper.
We used observations at the study site in the following ways. First, we perform a limited calibration to the
ecohydrologic component of our model using observed ecohydrologic data at the study site. Besides this
limited calibration, no other model calibration is performed. Then, with the calibrated model we test
whether realistic frequency and magnitude of runoff events can be generated consistent with other
regional observations. Finally, at the end of LEM experiments, modeled landscapes are qualitatively com-
pared against the observed topography of the field catchment to elucidate the differences between the
morphologies of the opposing NFS and SFS.
3.1. Ecohydrologic Model Confirmation
We confirm the ecohydrology component of the model against field observations at the McKenzie Flats area, in
the north-eastern quadrant of the SNWR. At the site, vegetation is predominantly composed of warm season C4
grassspecies,suchasblackgrama(Bouteloua eriopoda)andbluegrama(Bouteloua gracilis), and soil is loamy
sand [Moore,2012].Despiteitsflattopography,thissiteispreferred for model confirmation because of the avail-
ability of detailed climate, soil moisture, and flux observations. Ecohydrologic data used in this study include:
hourly precipitation, temperature, wind speed, and relative humidity between 1990 and 2009 at the Deep Well
meteorological station [Moore, 2012]; evapotranspiration fluxes between 1996 and 1999 from a Bowen Ratio
Energy Balance (BREB) tower located near the Deep Well station [Gosz, 2012a]; and soil moisture content of the
top 30 cm (i.e., average root depth for grass at the site [Kurc and Small, 2004]) measured using vertically installed
TDR (Time-Domain Reflectometry) probes at three different pits adjacent to the meteorological and BREB flux
measurements [Gosz, 2012b].
Figure 4. (a) Maps of the state of New Mexico (left), and the Sevilleta LTER (right) with the boundaries of the major ecosystem types, loca-
tion of the study catchment, and a weather station used to characterize climate forcing; (b) 2 m aerial orthophoto of the study catchment
used to confirm model simulations; (c) and (d) close-up photographs of segments of NFS and SFS, respectively, showing differences in veg-
etation and hillslope morphology on opposing aspects.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1136
MODIS LAI product is used to
confirm model predictions of
live LAI. MODIS LAI is available
at 1 km spatial resolution with
8 day intervals since 2000. A
window of 3 by 3 MODIS grid
cells overlaying the Deep Well
site at the center are obtained
from NASA Land Processes Dis-
tributed Archive Center (LP
DAAC), and the spatial mean of
LAI from the nine grid cells are
calculated to compare against
model predictions.
Concurrently measured BREB-
derived ET
a
and soil moisture
can be used to estimate the
physical soil parameters (s
wp
and s*) of the ET
a
model [Kurc and Small, 2004; Chen et al., 2008]. The loamy
sand soil type is defined as Berino soil in the metadata of the soil moisture observations by Gosz [2012b].
The bulk density of Berino soil is given as 1.54 g/cm
3
[Soil Survey Staff, 2012]. The soil porosity, n, using the
particle density of mineral soil (2.65 g/cm
3
), is estimated as n50.42. The ET
a
-s relationship (equation (10))
based on the piecewise linear dependency of ET
a
to sis fitted visually to the ET
a
-sdata pairs, as in Vivoni
et al. [2008]. For n50.42, the corresponding values of s
wp
and s* found are 0.17 and 0.31, respectively. These
values are within the range of s
wp
and s* reported in the literature for loamy sand soil [Laio et al., 2001].
A small flat surface represented with a mesh of 20 nodes is generated to test the proposed ecohydrologic
component of the model. CHILD is run for the period 1990–2009 driven by daily meteorological data
obtained by averaging hourly measurements. Table 3 reports the model parameters used in this simulation,
which are selected with limited calibration of the ecohydrological component of the model. Soil parameters
are first obtained from the cited literature only on the basis of soil textural type and soil moisture observa-
tions (s
wp
and s*). Experimenting with the model with fixed soil parameters, we identified GT,DT,WUE,k
sl
,
and k
dd
as the most critical parameters that control biomass dynamics in the model. GT and DT designate
the onset of growth and dormancy periods, respectively. WUE defines the amplitude of LAI during the grow-
ing season, while k
sl
and k
dd
control the decay rate of live and dead biomass. The values for WUE,k
sl
and k
dd
are selected by varying them within the ranges reported in the literature. For example, Larcher [1995]
reported WUE for C4 plants between 0.003 and 0.005 (kg DM/kg H
2
O). Monson et al. [1986] measured WUE
for blue grama in the 0.0056–0.012 (kg CO
2
/kg H
2
O) range. The value used for k
sl
is taken from Istanbulluo-
glu et al. [2012], while k
dd
is calibrated.
Figure 6 plots the ecohydrological response of the CHILD model driven by observed meteorological data at
the Deep Well site. Modeled soil moisture is in overall good agreement with observations in the 1996–2006
period (Figure 6b), capturing the magnitudes of soil moisture pulses and the shapes of soil moisture decays
reasonably well (daily Nash-Sutcliffe efficiency, NSE50.76). The model, however, slightly overestimates the
soil moisture in the winters of some years (e.g., 2005). This may be attributed to winter greening of some C3
species in the field, lowering the observed soil moisture through transpiration, while the single species veg-
etation used in the model would not sufficiently capture this early growth. This argument is supported by
the MODIS-LAI data, which shows early growth in 2005 while the modeled growth is delayed (Figure 6d).
Despite its simplicity, the model generally captures the observed peaks in ET
a
during the rainy season (Fig-
ure 6c). Most notably, the predictive skill of the model is improved in the wet season of 1999, when a collec-
tion of relatively large storm pulses fell during a shorter wet season. With respect to the LAI, the model
captures the onset of greening, the magnitude of the peak LAI, and the onset of senescence of grass vege-
tation as compared with the MODIS- LAI data (Figure 6d). It can be noted that MODIS-LAI does not capture
vegetation dynamics when LAI goes below 0.2. This discrepancy may be attributed to the insensitivity of
MODIS-LAI data to variations around very small values. Technically MODIS-LAI product uses 0.1 as the lower
LAI limit for vegetated surfaces [e.g., Fang and Liang, 2005; Shabanov et al., 2005].
Figure 5. The slope-upslope catchment area (S-A) plot of northern and southern aspects of
the catchment indicated by a star in Figure 4a. The plot is produced by binning and averag-
ing of local slopes observed within a given upslope catchment area range, and plotting the
binned slope data with respect to area.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1137
With this confirmation of the model at the Deep Well site, the role of aspect on modeled vegetation produc-
tivity is demonstrated by running CHILD at a small headwater catchment (0.22 km
2
) in the SNWR study
basin (Figure 4b, location indicated by star), using the Deep Well climate data. Modeled live biomass cover
(g DM/m
2
) is presented in two snapshots of the model corresponding to before (end of May) and after (late
August) the NAM of 1998 (Figures 7a and 7b, respectively). Modeled time series of mean spatial soil mois-
ture content (m
3
/m
3
) for 1997–2000 are plotted in Figure 7c. In both cases, NFS hold more grass biomass
than SFS as a result of greater soil moisture accumulation during the winter (Figure 7c). During the summer,
similar soil moisture contents are modeled on the opposing hillslope aspects. Given the NAM is the major
erosive flood-generating season in the region [e.g., Guti
errez-Jurado et al., 2007], it is anticipated that the dif-
ferences in the modeled vegetation biomass between the NFS and SFS would influence catchment evolu-
tion over geomorphic time scales. While the single plant functional type assumption in the current model
limits the application of the model at the SNWR study site, where juniper trees and grasses with higher total
biomass coexist on NFS and shrubs dominate SFS, the current model is deemed appropriate to examine the
ecohydrologic ramifications of aspect on geomorphic evolution as the model gives more plant biomass on
NFS, consistent with observations, than on SFS.
Table 3. Model Parameters Definitions and Sources for the Values Used in the Model
Parameter Definition Variable Value Method
Geomorphology Parameters
Mean grain size diameter d
50
0.007 m Approximated
Manning’s roughness for bare soil n
s
0.05 (2)Engman [1986]
Manning’s roughness for reference vegetation cover n
v
0.5 (2)Engman [1986]
Reference vegetation cover V
R
0.95 (2)Istanbulluoglu and Bras [2005]
Reference vegetation cover exponent x0.5 (2)Istanbulluoglu and Bras [2005]
Hillslope diffusivity coefficient K
d
0.003 m
2
y
21
Hurst et al. [2013]
Hillslope diffusivity critical gradient S
c
1.2 (2)Roering et al. [1999, 2007]
Sediment transport coefficient j20 (2)Simons and S¸ent
urk [1992]
Critical shear stress for sediment transport s
c
5 Pa Estimated
Detachment2limited exponent p
d
2.377 (2)Nearing et al. [1999]
Transport2limited exponent p
f
2.5 (2)Govers [1992]
Detachment2limited erodibility coefficient k
d
0.449 m Pa
22.377
y
21
Nearing et al. [1999]
Soil Parameters
Porosity n0.42 (2) Estimated
Bare soil infiltration capacity I
c,bare
12 mm h
21
Approximated
Vegetated infiltration capacity I
c,veg
36 mm h
21
Approximated
Anisotropy ratio A
R
1000 (2) Approximated
Empirical parameter in the Campbell model b4.85 (2)Laio et al. [2001]
Hygroscopic capacity s
h
0.1 (2)Laio et al. [2001]
Wilting point s
w
0.17 (2) Estimated
Incipient stomata closure s* 0.31 (2) Estimated
Field capacity s
fc
0.52 (2)Laio et al. [2001]
Climate Parameters
Mean annual precipitation MAP 249.6 mm y
21
Observation
Fraction of wet season duration F
wet
0.252 (2) Observation
Fraction of wet season precipitation F
p
0.496 (2) Observation
Poisson storm scale factor a6.0 (2) Calibrated
Rainfall intensity after scale calibration (wet/dry seasons) p08.73/2.91 mm h
21
Small [2005]
Interstorm duration after scale calibration (wet/dry) T0b114.3/253.3 h
21
Small [2005]
Storm duration after scale calibration (wet/dry) T0r0.77/1.71 h
21
Small [2005]
Vegetation Parameters
Effective rooting depth D
r
0.30 m Kurc and Small [2004]
Bare soil evaporation coefficient k
s
0.7 (2)Istanbulluoglu et al. [2012]
Growth threshold GT 4.93 mm d
21
Calibrated
Dormancy threshold DT 4.93 mm d
21
Calibrated
Green biomass senescence coefficient k
sl
0.012 d
21
Istanbulluoglu et al. [2012]
Dead biomass decay coefficient k
dd
0.05 d
21
Calibrated
Constant for dead biomass loss adjustment T
dmax
6.0 mm d
21
Calibrated
Water use efficiency WUE 0.005 kg CO
2
/kg H
2
O Approximated
Conversion factor of CO
2
to dry biomass x0.55 kg DM/kg CO
2
Scholes and Walker [1993]
Drought2induced foliage loss factor k
sf
0.04 d
21
Ivanov et al. [2008a]
Specific leaf area for live biomass c
g
4.7 m
2
kg
21
DM Istanbulluoglu et al. [2012]
Specific leaf area for dead biomass c
d
9.0 m
2
kg
21
DM Istanbulluoglu et al. [2012]
Maximum leaf area index LAI
Max
5.0 (2)Istanbulluoglu et al. [2012]
Vegetation destruction parameter k
v
100 Pa
21
y
21
Calibrated
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1138
3.2. Annual Runoff and Flood Characteristics
In this section, we evaluate runoff coefficients, RC (percent of the ratio of mean annual runoff, MAR,to
mean annual precipitation, MAP, or 100*MAR/MAP), and flood return periods simulated by CHILD, with
those reported in the literature for the semiarid southwest US and more specifically in central New Mex-
ico. To evaluate the runoff component, the model is run with calibrated soil and vegetation parameters
on an evolved catchment, forced by generated climate for 10,000 years. Runoff from each simulated
storm is recorded. For this analysis, we run the model 5 times, using MAP from 200 to 400 mm with
50 mm increments to examine the overall model response for semiarid conditions. In each run, the wet
season (July, August, and September) brings 50% of the MAP, as observed in the region during the
NAM. The modeled runoff coefficients increased from 3.1% (MAP5200 mm) up to 4.0%
(MAP5400 mm).
At the study catchment, Guti
errez-Jurado et al. [2013] reported measured RCs in field plots (4 m by 2 m) on
NFS and SFS as 0.1% and 7%, respectively, which clearly reflects the fundamental role of aspect on runoff gen-
eration even at the plot scale. For a small catchment made up equally of north and south aspects, this would
roughly give a 3.5% RC. At the scale of the Rio Puerco (160,000 km
2
) and its two major tributaries, Moln
ar and
Ram
ırez [2001] reported RCs between 0.89% and 2.96%. While their values are lower than our simulations, it is
highly likely that lower RCs observed in their study can be due to transmission losses in channels and reduced
amounts of synchronous runoff-producing storms in larger basins (e.g., Dunkerley, 1992; Mudd, 2006; Parsons
et al., 1999; Vivoni et al., 2006). In an earlier study, Drissel and Osborn [1968] reported a RC of 3.3%, in a
173 km
2
basin located in east central New Mexico, consistent with our model runs.
On the wetter end of the simulated climate range, CHILD simulations agree with the RCs observed in vari-
ous subbasins of the Walnut Gulch Experimental Watershed (WGEW) in southeastern Arizona, with similar
climatology (MAP5312 mm) and coarse soil texture [Goodrich et al., 2008]. In the WGEW, Stone et al. [2008]
reported RCs of 1.9% and 4.5% for experimental catchments with areas of 8.98 km
2
and 0.06 km
2
, respec-
tively. Stone et al. [2008] also attributed the observed reduction in the RC with catchment area to growing
channel losses.
Figure 6. Confirmation of the ecohydrology component of the CHILD LEM at the Deep Well site, Sevilleta National Wildlife Refuge. Figures plot time series of: (a) observed precipitation;
(b) modeled and measured (in three different soil pits) depth-averaged soil moisture (s) in the 30 cm root zone; (c) modeled and Bowen ratio-estimated evapotranspiration (ET
a
); and
(d) modeled live leaf area index (LAI) and MODIS-derived LAI.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1139
We compare the return periods of modeled annual maximum daily runoff for MAP5250 mm with measured
runoff in flume studies of relatively small-scale basins (<10 km
2
) at the WGEW. Other existing data for the
central New Mexico region are often for much larger basins. In headwater catchments, most erosive floods
are generated by localized high-magnitude storm pulses. Although the WGEW catchments receive higher
MAP (312 mm), they also experience a longer NAM [Goodrich et al., 2008], making storm events comparable
with those in central NM. Therefore, we anticipate relatively similar flood generating conditions in both NM
and AZ sites. Figure 8 shows the observed and modeled annual maximum daily runoff and their return peri-
ods. In the WGEW, the peak flow data were collected from three flumes that drain small catchments of vary-
ing sizes for 54 years [Stone et al., 2008]. Despite CHILD predicts slightly larger flood magnitudes for return
periods of 10 or less years, modeled flood magnitudes are particularly consistent with flume 4 and flume 125
of WGEW data for higher return periods. Flume 3 with a larger drainage area produces lesser runoff argu-
ably as a result of channel losses and spatial variability of storm intensities. The analysis presented in this
section demonstrates the credibility of the model in generating storm runoff at magnitudes commensurate
with small semiarid basins in the southwest U.S.
4. Results
We conducted simulations using spatially distributed radiation (Rad-Spatial) and spatially uniform radiation
(Rad-Uniform) with CHILD LEM to study the role of solar radiation on landscape and biomass development.
Model parameter values are selected to represent the environmental conditions in the central New Mexico
catchment where aspect-control on ecogeomorphic organization is observed.
Figure 7. Modeled live biomass cover (g DM/m
2
) at a small headwater valley in the SNWR catchment shown in Figure 4b (location indicated by a star) for: (a) pre-monsoon, end of May
1998; and (b) end of monsoon, late August 1998 conditions; (c) time series of modeled mean-spatial volumetric soil-moisture content hsiin the root zone at this small headwater valley
for the 1997–2000 period. Dashed lines in Figure 7c refer to field capacity, s
fc
, incipient stomata closure, s*, and plant wilting point, s
wp
thresholds.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1140
The Deep Well site has limited (19
years) of meteorological data to
parameterize the PRP rainfall
model. Hence, seasonal storm
characteristics p;Tr;Tb

are esti-
mated from empirical relations for
rainfall climatology developed for
the semiarid southwest U.S. as a
function of MAP, wet (F
wet
), and
dry (F
dry
) season precipitation
expressed as fractions of MAP, and
the start and end days of the NAM
[Small, 2005; Istanbulluoglu and
Bras, 2006]. Table 3 shows the esti-
mated storm intensity, duration,
and time between storms for MAP
of 250 mm and calculated F
wet
and
F
dry
from the Deep Well site.
PET forcing of the model is obtained
from a sinusoidal function of TF
max for reference grass as a function of DOY. The sinusoidal function is fitted to
the 19 year long calculated daily TF
max data for the Deep Well site to represent the seasonal changes in PET forc-
ing. In the Rad-Uniform simulations, solar radiation is assumed to be spatially uniform over the simulated
domain, and TF
max is used directly from the sinusoidal function. In the Rad-Spatial simulations, the clear sky radia-
tion is calculated throughout the year as a function of latitude, local slope, and aspect (see Appendix A), and
used in equation (12a) to calculate the clear-sky ratio (R
solar
). For hillslopes, TF
max is scaled with R
solar
,tocalculate
TS
max for a given slope and aspect of a landscape element (equation (12a)).
The initial domain used identically in all the model simulations is a 900 m by 900 m inclined (7%-grade)
east-facing elevation field constructed of Voronoi polygons with 20 m node spacing and added random ele-
vation perturbations with a mean of 2.5 m. Drainage is only permitted on the bottom of the sloping side of
the domain. The role of solar radiation on landscape and plant biomass development is examined under
the conditions of no uplift, a low rate of uplift (0.05 mm/yr), and a high rate of uplift (0.10 mm/yr), for both
Rad-Spatial and Rad-Uniform conditions. These uplift rates are within the ranges of the long-term (640
ka) average incision [Dethier, 2001], and denudation rate estimates in the region [Bierman et al., 2005; Clapp
et al., 2001]. Because our focus is to examine and isolate the role of solar radiation on landscape evolution
in this study, we deliberately choose to exclude any observed spatial patterns in soil and vegetation proper-
ties as input to the model experiments, and use the same rainfall time series as model forcing in all the
numerical experiments.
In the model, we assume that the landscape is soil-mantled (no bedrock control), and transport-limited. This
assumption is consistent with the regional geology of alluvial fan deposits of the Plio-Pleistocene epoch
and our field observations where we observed alluvial sediment deposits in valleys and absence of any rock
outcrops on hillslopes. Soil texture for water balance modeling is assumed homogenously loamy sand.
Coarse soil texture with high gravel contents covers hillslopes in the study site. McMahon [1998] reports
approximately 60% and 35% gravel content in the surface substrate of the SFS and NFS, respectively. Hence,
a fine gravel surface substrate with the mean grain size diameter, d
50
57 mm is used for modeling sediment
transport capacity. The critical shear stress, s
c
, for d
50
57mmiss
c
55 Pa according to the Shields equation
[Shields, 1936], used for both detachment and transport capacity equations. The transport coefficient of the
latter is calculated with d
50
57 mm to be consistent with the selected s
c
(Table 3).
We took the exponents of the sediment transport (p
f
) and detachment capacity (p
d
) models from Govers
[1992] and Nearing et al. [1999], respectively. Istanbulluoglu et al [2003] compared sediment yield estimates
from incised gullies on steep slopes against predictions using the Govers [1992] equation and found the
equation suitable for natural landscapes. Nearing et al. [1999] conducted experiments to relate net detach-
ment rate with shear stress in Walnut Gulch basin where coarse sediments dominate the surface substrate
Figure 8. Comparison of the annual maximum daily runoff (mm/d)—return period rela-
tion between the CHILD model simulations and observations from several small water-
sheds at the Walnut Gulch Experimental Watershed, Tombstone, AZ.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1141
similar to our catchment [Nearing et al., 1999]. Vegetation erodibility is calibrated such that with major
floods vegetation in the main channels are disrupted significantly.
Colluvial transport is modeled with the nonlinear hillslope diffusion rule [Roering et al., 1999]. A general
increase in hillslope diffusivity with MAP, climate seasonality, and ecosystem type has been reported in the
literature [Hanks, 2000; Hurst et al., 2013]. Higher long-term average K
d
values were reported for forested
basins in humid climates undergoing bioturbation [Black and Montgomery, 1991; Nash, 1980] than arid and
semiarid grass/shrubland ecosystems where physical disturbances govern transport [Hanks, 2000; Gabet,
2003]. Significant changes in hillslope diffusion were reported with ecosystem change from grass to shrubs
vegetation types [Hughes et al., 2009]. Our current model, however, only tracks grass biomass in space and
time. While there is consensus that grass cover shields the soil surface against rain splash, bioturbation
activity may dominate diffusive fluxes as grass biomass increases. We could not find any strong evidence to
support a conceptual model to relate hillslope diffusion coefficient to grass biomass in a semiarid climate
regime. Hence, a constant hillslope diffusivity value is used in the model.
4.1. Landscape Morphology
Model experiments are run for 800,000 years until approximate uplift-erosion equilibrium is reached. Figure
9 shows the time series of mean elevation of the simulated landscapes at 5,000 year resolution. Figure 10
presents the plan views of modeled topographies colored with respect to the size of the drainage area of
each cell. Figure 11 illustrates the 3-D features of the modeled landscapes forced with high uplift rate.
Uplift raises the mean elevation of the modeled landscapes, and leads to a clear separation among simula-
tions (Figure 9). In absence of uplift, topography decays with a diminishing rate over time, as landscape ele-
vations drop under continuing erosion. With uplift, the mean elevation of the basin attains an approximate
constant (with some fluctuations), illustrating the dynamic uplift-erosion balance in the basin. Rad-Spatial
simulations maintain slightly higher mean elevations than Rad-Uniform simulations. At the end of the simu-
lations, elevation differences between Rad-Spatial and Rad-Uniform cases grew with uplift from 0.30 m
under no uplift, to 1.37 m with high uplift, corresponding to 3.29% and 5.18% of the mean elevations of
their respective Rad-Spatial landscapes. Higher mean elevations under Rad-Spatial suggest overall steeper
slopes across the modeled domains as in all simulations the elevation of the outlet point is fixed in space
and time, which leads to increased landscape relief as elevations grow. These findings imply: (a) spatially
explicit solar radiation leads to lower erosion potentials (for a given slope and drainage area) across the
landscape, and therefore landscapes adjust to occupy higher mean elevations (and steeper slopes) to main-
tain erosion-uplift balance; and (b) the influence of solar radiation in the evolution of modeled catchments
intensifies by uplift.
We use drainage density, D
d
, and hillslope asymmetry, HA, indices to provide metrics as the basis of objec-
tive comparisons between the modeled and actual field catchments (Table 4). D
d
is the ratio of the total
length of channels to catchment area [e.g., Tarboton et al., 1991]. A channel network is extracted using a
constant support area threshold of 2,400 m
2
, approximate inflection point of the S-Aplot (Figure 5). Follow-
ing Poulos et al. [2012], the north-south HA,HA
N-S
, is estimated as the log
10
of the ratio of median slopes of
the northern and southern aspects (Table 4). HA
N-S
>0 reported for modeled catchments in Table 4 indicates
steeper NFS than SFS. To confirm this finding, we test the null hypothesis that mean slope values of NFS
and SFS in the modeled catchments are identical. With the calculated pvalue (p<0.001), this hypothesis is
rejected for all the experiments, suggesting that slopes of NFS and SFS are statistically different regardless
of the rate of uplift. Regional DEM analysis found NFS steeper than SFS for both a50.05 and a50.01 levels
[Istanbulluoglu et al., 2008]. A consistent pattern in the calculated metrics is that both D
d
and HA
N-S
are
higher in the Rad-Spatial simulations and increase with uplift. For example, HA
N-S
of 0.1 indicates that NFS
are 26% steeper than SFS in the basin. D
d
and HA
N-S
values of the field-site catchment fall between the no
uplift and low uplift simulations. HA
N-S
can be a function of different factors including uplift rate, geology,
vegetation, and climate. For example, Poulos et al. [2012] estimate HA
N-S
from 30 m DEMs for the Gabillan
Mesa in the central California Coast Ranges located at the 36N latitude and Dry Creek located at the 44N
latitude within the 0.10–0.15 range.
The differences in the metrics discussed above are manifested in the modeled landscape structure (Figures
10 and 11). In the no uplift experiments, modeling spatially varying radiation leads to the development of
linear and more closely spaced channels (Figures 10a and 10b). In the Rad-Uniform case, upwardly
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1142
migrating channels branch into both north and south facing aspects (see blue arrows), developing wider
headwater valleys (Figure 10a), and capturing surface runoff into three major dendritic channels. In the Rad-
Spatial case, however, northward channel development is inhibited by vegetation, as in the model vegeta-
tion grows more densely on NFS than SFS (e.g., Figure 7). This results in the development of narrower val-
leys with more linear channels fully advanced to headwater slopes.
Uplift intensifies the role of solar radiation on modeled landscapes. In the Rad-Spatial simulations, channels
more extensively develop on SFS with increasing uplift, as can be profoundly observed in the first valley on
the left (see blue arrow) in Figures 10b, 10d, and 10f. The northward expansions of channel heads are
responsible for the development of hillslope asymmetry. Figure 11 more clearly demonstrates the differen-
ces in the development of landscape elevations and valley networks between the Rad-Uniform and Rad-
Spatial simulations. Figure 11b shows how the northward expansion of the channels of the first valley into
SFS (see the first left red arrow) inhibits the development of the adjacent valley (see middle red arrow) by
capturing its drainage area. The last arrow on the right points at a valley that fully developed on the Rad-
Spatial case (but not in Figure 11a) when channels in the adjacent valley could not incise on NFS, as they
did in case of the last valley on the right in the Rad-Uniform simulation (Figure 11a).
4.2. Slope-Area and Vegetation-Area Relations
Contributing drainage area, A, is a measure of water supply at a point, used for both fluvial transport and
plant growth, while slope and aspect regulate the amount of local water loss to ET. We plot local values of
slope, S, and vegetation cover fraction, V
t
, with A, grouped with respect to north (315–45) and south (135–
225) aspects for all simulations to examine the associations between modeled vegetation cover and land-
scape morphology (Figure 12). For the S-A plot, the elevation field at the end of each model run is used,
and Svalues for each aspect are binned and averaged with respect to given ranges of A. To capture the
temporal dynamics of V
t
,V
t
for each model element is stored after every storm event in the last 100 years of
the simulations, and the temporal average of V
t
,Vt, for 100 year is plotted as a function of Ain Figure 12.
In the Rad-Spatial simulations, modeled NFS are steeper than SFS for A<10
4
m
2
(Figures 12a, 12c, and
12e), consistent with the observed S-A plot of the field catchment (Figure 5). The S-A data of the Rad-
Uniform simulations plot between those of the NFS and SFS of the Rad-Spatial simulations. Landscapes get
steeper, and the separations between the slopes of opposing aspects become more pronounced as uplift
Figure 9. Time series of the mean elevation of modeled landscapes at a 5,000 year resolution under spatially uniform (Rad-Uniform) and spatially variable (Rad-Spatial) radiation condi-
tions for 0.00 mm/y, 0.05 mm/y (low), and 0.1 mm/y (high) uplift rates.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1143
increases. The coefficient of variation (CV
s
) calculated for each slope bin show lower slope variability for hill-
slopes (CV
s
50.1–0.45) than channels (CV
s
50.63–0.82). Lower values in the respective CV
s
ranges correspond
to uplift simulations. While the influence of aspect is subtle on CV
s
, NFS exhibit slightly more variable slopes
in all the simulations.
Figure 10. Plan views of numerical experiments for Rad-Uniform (left) and Rad-Spatial (right) conditions under 0.00 mm/y, 0.05 mm/y, and 0.1 mm/y uplift rates (top to bottom). Arrows
are explained in the text.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1144
The S-A and Vt2Arelations are related (Figure 12). There is clearly a contrasting behavior of vegetation
response with Aon opposing slopes. Vton NFS first shows a slight positive dependence to Aon hillslopes
(e.g. A<2x10
3
m
2
), attains a local maximum, and drops with further increase in Ato a nearly constant
value in channels near the outlet. This response is in-phase with the S-A relation. NFS (steeper than 10)
have significantly lower R
solar
, especially during autumn and winter (Figure 3a), that reduces T
max
and
enhances soil moisture in the beginning of the growing season (Figure 7c). Wetter soils leads to lower plant
water stress, supporting denser grass cover throughout the year. Loss of vegetation on NFS as Sdecreases
and Agrows suggests that an increase in upland moisture input with larger Acannot compensate the grow-
ing ET losses as a function of an increase in R
solar
as Sgets shallower. This leads to foliage loss as a result of
increasing plant water stress on flatter topography.
In contrast to NFS, steep SFS receive the highest R
solar
(Figure 3b) during autumn and winter, which ampli-
fies evaporation losses until the start of the growing season compared to NFS (Figure 7). Drier soil moisture
conditions in the beginning of the growing season and slightly elevated R
solar
on SFS during the growing
season hamper vegetation productivity, leading to a marked difference between the Vt2Arelations of NFS
and SFS for A<10
4
m
2
. Drop in local Sas Agets large in concave south-facing valleys improves soil mois-
ture as a result of larger Aand reduced ET, supporting more productive vegetation growth (i.e., positive rela-
tionship between V
t
and A) on southern aspects of moderately sloped valleys (Figures 12b, 12d, and 12f).
Model results related to vegetation variability can be confirmed with some limited empirical data. Flores Cer-
vantes et al. [2014] investigated associations between topography and grassland biomass in the Walnut
Gulch Experimental Watershed, AZ and in central NM using Landsat NDVI data. As in our study, they found
a strong impact of aspect control on biomass production on NFS. On those slopes, greater biomass cover
was observed in landscape positions with intermediate values of upslope drainage areas with steep slopes.
Biomass was found to increase more persistently with drainage area on SFS. Their observations conceptually
support our model results presented in Figures (12b, 12d, 12f; and 14).
Examining the responses of S-A and Vt2Arelations to increase in uplift in Figure 12 further reveals a solar
radiation-induced positive (negative) feedbacks between slope evolution and vegetation on NFS (SFS).
Higher rates of uplift drive the opposing hillslopes to get steeper as the landscape evolves toward an uplift-
Figure 11. Elevation map (m) of modeled topography driven by high uplift (0.1 mm/y) for: (a) Rad-Uniform; and (b) Rad-Spatial cases. Channels are mapped using a 2,400 m
2
support
area threshold. Red arrows from left to right indicate the formation of channels on south-facing aspects, inhabitation of channel development in the adjacent valley, and a fully devel-
oped valley at northing5700m which did not develop under Rad-Uniform conditions.
Table 4. Drainage Density, D
d
(km/km
2
), and North-South Hillslope Asymmetry, HA
N-S
, Indices for Modeled and Field Catchments
Parameter Field Site Radiation No Uplift Low Uplift High Uplift
D
d
(km/km
2
) 10.94 Spatial 9.43 11.00 11.39
Uniform 9.87 10.33 10.61
HA
N-S
0.086 Spatial 0.060 0.098 0.131
Uniform 0.037 20.0002 0.002
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1145
erosion equilibrium. Slope steepness has a direct impact on solar radiation but in opposite directions in the
northern and southern aspects. Decreasing R
solar
as NFS get steeper with growing uplift result in wetter NFS
that can support denser vegetation cover for a given A. On the contrary, increasing R
solar
as SFS get steeper
result in drier SFS, that support sparser vegetation cover for a given A(Figures 12b, 12d, and 12f). In the
model, vegetation inhibits the effectiveness of flow shear stress and reduces rates of erosion and sediment
transport (Figure 2). As a result of this ecohydrologic feedback to geomorphic transport, an uplift-erosion
balance at both local and catchment scales (as imposed by the conservation law) can only be achieved in
the model by developing steeper NFS and shallower SFS than the Rad-Uniform simulations. Increase in
uplift further amplifies this feedback and differences between S-A and Vt2Arelations of NFS and SFS grow.
In this model, we neglect differences in plant functional types between the opposing slopes as only grass
vegetation type was used. However, if instead of grasses, a tree-grass savanna ecosystem was used in the
model for NFS, a higher hillslope diffusion consistent with the literature would have been preferred. An
assumption of increased hillslope diffusivity with vegetation biomass in our model would reduce slopes in
the diffusion-dominated portions of NFS (area<350m
2
in Figure 5) and shift the location of the valley
head on the S-A plot to a larger drainage area. However hillslope diffusion does not have a strong signature
in the modeled landscapes as 20 m node spacing was used in the initial domain for computational effi-
ciency. As a result, diffusion-dominated hillslopes are only confined to the first two bins of the S-A data (see
uplift runs Figures 12c and 12e).
4.3. Spatiotemporal Soil Moisture and Vegetation Dynamics
Understanding and representing the spatial and temporal patterns of soil moisture is critical in the coupled
modeling of ecohydrologic and geomorphic processes because of the importance of soil moisture in the
generation of flood response and ecosystem processes [Guti
errez-Jurado et al., 2007; Vivoni et al., 2009;
Penna et al., 2011]. A number of studies have demonstrated that the coefficient of spatial variation of soil
moisture h(volumetric water content), CVh, varies as a function of its spatial mean mean h,hhi. It has been
shown that CVhattains a maximum within an intermediate hhistate of the field, and drops toward its lowest
values as the watershed dries and wets, producing a convex parabolic CVh2hhirelationship [Western et al.,
2003; Ryu and Famiglietti, 2005; Choi and Jacobs, 2007; Famiglietti et al., 2008; Mascaro et al., 2011]. When
Figure 12. Slope-area, S-A, (left) and vegetation-area, Vt2A, (right) relations for NFS and SFS of the Rad-Spatial simulation, and for the whole landscape of the Rad-Uniform simulation
for: (a, b) no-uplift; (c, d) low-uplift (U50.05 mm/y); and (e, f) high-uplift (U50.1 mm/y) model experiments.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1146
the soil moisture data are plotted in the CVh2hhiphase space as a function of time (i.e., a trajectory), how-
ever, a hysteretic pattern emerges, in which the spatial variability of soil moisture during wetting and drying
phases follows different trajectories [Teuling et al., 2007; Ivanov et al., 2010; Vivoni et al., 2010; Rosenbaum
et al., 2012].
We output fields of root-zone average soil moisture and vegetation cover following each storm in the last
100 years of the model simulations with high uplift, and calculate the mean spatial soil moisture and vege-
tation cover hhi;hVtiðÞand their coefficients of spatial variations CVh;CVVt
ðÞafter each storm event. Figures
13 and 14 present soil moisture CVh2hhiand vegetation CVVt 2hVtidata pairs for both Rad-Spatial and Rad-
Uniform scenarios, along with the maps of selected soil moisture and vegetation states at the high and low
ends of their modeled ranges.
The CVh2hhirelation in Figure 13a is bounded by low CVhvalues on both dry and wet ends of hhi. On the
dry-end, hhivalues are generally within the range of hygroscopic water content and plant wilting point
(0.04–0.07), and on the wet-end hhivalues are between field capacity and porosity. The spatial variability
of soil moisture shows fundamental differences between the Rad-Spatial and Rad-Uniform simulations. The
CVh2hhidata pairs of the Rad-Uniform scenario establish a lower limit in the CVh2hhispace, with a very
slight increase (decrease) in spatial variability as basin wets (dries). In the Rad-Spatial simulation, CVhgener-
ally maximizes about the mid value 0.12 of hhi(e.g., point C) and decreases toward the wet and dry ends.
Distinct differences between the two radiation scenarios suggest that the spatial variability of solar radiation
drives the convex parabolic CVh2hhirelation, consistent with the generally reported patterns in the litera-
ture that used field observations [Ryu and Famiglietti, 2005; Choi and Jacobs, 2007] and numerical modeling
[Lawrence and Hornberger, 2007; Vereecken et al., 2007; Ivanov et al., 2010].
We plot trajectories of CVhas a function of hhifor 2 years of soil moisture wetting-drying cycles to further
examine the physical processes that underlie the spatial and temporal evolution of soil moisture (Figure
13a). For a clear illustration of the annual CVh2hhirelationship, we start the trajectories with the driest soil
moisture condition before the onset of the wet season (i.e., NAM) in July, and end when the minimum soil
moisture is reached in the following year prior to NAM. The size of each loop in the figure is characterized
by the amount of wet season rainfall. The outer (inner) loop corresponds to a relatively wetter (drier) years
with 317 mm (160 mm) of total rainfall (from July to July). To investigate the topographic controls on the C
Vhstates, the simulated maps of spatial soil-moisture distribution for the wetter year are shown for the dri-
est and wettest hhiconditions (Aand Bin Figure 13a), and the highest CV point (Cin Figure 13a) in Figures
13b, 13c, and 13d.
With the driest mean soil moisture of the year (point A, Figure 13a), both aspects are under the plant wilting
point, while NFS have slightly greater soil moisture contents (aspect-controlled pattern) than SFS (Figures
13b). The onset of NAM wets the topography through several successive storms and hhievolves to its high-
est value of the year during NAM (point B, Figure 13a). Interestingly, during this wetting process, CVhfirst
shows a slight increase with hhi, but then drops with the final storm event as hhitakes its maximum value
by mid-September of the simulated year. A similar change in the sign of the CVh2hhirelation can be seen
in the Rad-Uniform simulation as hhigets larger than 0.23, a value slightly higher than the field capacity
of the soil (h
fc
50.22). This model response is a result of a growing influence of lateral soil-moisture redistrib-
ution over the domain, activated by the exceedance of the field capacity threshold. Enhanced hydrologic
network connectivity with lateral flow is shown to reduce the spatial variability of soil moisture (i.e., homog-
enizing effect) [Ivanov et al., 2010]. In the mapped soil moisture field for point Bof the CVh2hhirelation (Fig-
ure 13c), soil moisture is greater than field capacity across the domain, but shows a fairly narrow spatial
range (0.3–0.33). The control of topographic convergence (network-controlled pattern) can be clearly seen
on modeled soil moisture. Besides the role of lateral transport, lack of aspect influence on the incoming
solar radiation during NAM (see wetting range in Figure 3), arguably contributes to the low spatial variability
of soil moisture between points Aand Bin the CVh2hhidomain (Figure 13a) by imposing a relatively uni-
form T
max
across the modeled domain.
Consistent with Ivanov et al. [2010], we have identified two phases in the drying process of the catchment
(Figure 13a). Starting with a few storm events after the wettest point B, during phase-1 of drying, CVh
increases as hhidrops, until the peak CVhis reached at point C. In phase 2, CVhdrops rapidly with further
decrease of hhi, to the driest pre-NAM soil moisture state. This hysteretic pattern, when viewed in relation
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1147
to soil moisture thresholds, reveals the underlying processes that control soil moisture variability. Starting
from point B(Figure 13a), as soil moisture dries below the field capacity (h
fc
50.22) lateral transport ceases
gradually across the catchment. Without the homogenizing effect of lateral flow, the CVh2hhirelation
presents a negative linear response, carrying CVhto a maximum value (drying phase 1).
The phase 1 of drying corresponds to the fall-winter time frame (October–March) during which the modeled
region gradually grows a large contrast in the incoming radiation between the NFS and SFS (Figure 3). Differ-
ential drying leads to a strong aspect control on the modeled soil moisture (aspect-controlled pattern), taking
CVhto its highest value at point C. In the mapped soil moisture field for point C(Figure 13d), SFS experience
near wilting point conditions, while most NFS are above the threshold soil moisture for stomata closure
hs50:13ðÞand therefore lead to evapotranspiration at the potential rate. The drying phase 2 corresponds to
March–July period that triggers a rapid homogenization in the soil moisture field. During this period, the
incoming solar radiation equilibrates across aspects, and the growing season first begins on NFS where soil
moisture is sufficient for growth (Figure 7). The following dry NAM season also follows a similar wetting-
drying cycle (inner loop labeled Dryer Cycle), while the weakness of the NAM limits the size of the cycle.
The relationship between the coefficient of spatial variation of vegetation cover, CV
Vt
,andthemeanspatial
vegetation cover, hVti, in the Rad-Spatial simulation resembles an obtuse triangle for each year as illustrated
by the two continuous CVVt 2hVtitrajectories (Figure 14a). The outer loop is for the wettest year in the last
100 year period of the modeled data (538 mm, labeled Year 54), with the highest biomass production in the
model record. The inner loop is the CVVt2hVtiresponse of the same year for which the wetting and drying
phases of the CVh2hhitrajectories are presented in Figure 13, and will be discussed in this section. CVVt2hVti
trajectories start before the growing season with the lowest value of hVti(see Ain Figure 14a) of the year
Figure 13. (a) The coefficient of spatial variation of soil-moisture content in the root-zone, CVhplotted as a function of its spatial mean hhiover the modeled domain for Rad-Spatial and
Rad-Uniform simulations. Two annual CVh2hhitrajectories highlighted are from the Rad-Spatial scenario, for years with wetter (outer loop) and drier (inner loop) than average annual
precipitation. Modeled soil moisture is mapped on evolved topography for the wetter year (annual precipitation 317 mm) for days with: (b) the driest hhi; (c) wettest hhi; and (d) the
highest CVh. Note that the ranges of the color bars are different.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1148
typically in April or early May, and end with the last storm of the dry season in the following year (before the
start of NAM). Each point in the plot is plotted after a storm event.
Higher spatial variability of vegetation cover is observed when the mean biomass in the modeled domain is
low. Seasonal vegetation growth-decay dynamics, indicated by arrows on Figure 14a, lead to a hysteretic
response of CV
Vt
as a function of hVti. To illustrate this behavior, we map V
t
of points A,B, and Cof Figure
14a over the modeled elevation field (Figures 14b, 14c, and 14d).
NFS have relatively higher vegetation cover at point Awith minimum hVtibefore the beginning of the
growing season (end of April) (Figure 14b). This vegetation map corresponds to the early stages of the sec-
ond phase of drying (closer to point Cin Figure 13a). Soil moisture maps in Figures 13c and 13d show wet-
ter NFS throughout this drying phase. Starting from point Ain the CVVt2hVtidomain, the growing season
imposes a homogenizing effect in the modeled vegetation domain. Here, the growing season is identified
by general climatological conditions that lead to a net increase in biomass. CV
Vt
first falls rapidly during
growth until hVtireaches a value in the 0.1–0.2 range. This period occurs before the onset of NAM from end
of April to July, and is driven by the available root-zone soil moisture from the second phase of drying and
some sporadic rainfall pulses within the dry season. In some cases, excessive biomass that grows during
this ‘‘early growth’’ decays rapidly, forming a small growth-decay loop within the yearly evolution of the C
VVt2hVtidomain, as can be seen in the high biomass year.
As hVtiincreases during the modeled wet season (i.e., NAM), CV
Vt
remains at a relatively low and nearly con-
stant value. In Figure 14a, point Bmarks the end of the wet season (DOY 5281). The maximum value of hVti
in the CVVt2hVtidomain is reached soon after the wet season, stimulated by the wet-season soil moisture
Figure 14. (a) The coefficient of spatial variation of total vegetation cover fraction, CV
Vt
, plotted as a function of its spatial mean hVtiover the modeled domain for Rad-Spatial and Rad-
Uniform simulations. Two annual CVVt 2hVtitrajectories highlighted are from the Rad-Spatial scenario, for years with highest biomass (outer loop) and the wetter than average year
(inner loop) used in the soil moisture plots. Modeled vegetation cover is mapped on evolved topography for the wetter year (annual precipitation 317 mm) for days with: (b) the lowest
hVti; (c) high hVtishows south-facing slopes have denser canopy cover than north-facing slopes; and (d) the highest hVtiwith low CV
Vt
that shows north-facing slopes have denser can-
opy cover than south-facing slopes. Note that the range of the color bar in (b) is different than (c) and (d).
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1149
(DOY 5302 late October) (point C). During the entire growth period from points Ato Cin Figure 14a, the
pattern of soil moisture moves from an aspect-controlled pattern to a network-controlled pattern, starting
with moisture states closer to the mapped soil moisture state in Figure 13d, passing from Figure 13b, and
reaching the wettest point toward the end of the wet season in Figure 13c.
Figures 14c and 14d reveal an interesting switch in the control of aspect on modeled vegetation patterns,
which can be related to the spatial patterns of solar radiation and soil moisture. During the period when soil
moisture is sufficient to support vegetation growth (Figures 13c and 14c), typically very early in the spring,
SFS with slightly higher incoming solar radiation (Figure 3) show denser vegetation cover than NFS. How-
ever, this ‘‘early-greening’’ on SFS is relatively short-lived. At point Cwith the highest hVtiof the season
(DOY5302) (Figure 14a), there is a clear advantage of NFS and converging topography on vegetation pro-
ductivity (Figure 14d). From point Cin the CVVt 2hVtidomain, until the beginning of the growing season in
the forthcoming wet season, NFS will remain to hold more vegetation biomass, leading to growing CV over
time.
In their Landsat-derived observations, Flores Cervantes et al. [2014] found greater spatial variability of grass-
land productivity during the dry season (before NAM) compared to the wet season in central NM. They
reported CV50.7 (CV50.2) for the days with the lowest (highest) mean spatial grass biomass. The CV values
reported by Flores Cervantes et al. [2014] for the dry season are comparable with the modeled values plotted
in Figure 14a, while our model gives less variable vegetation cover during the wet season (Figure 14a). This
model limitation can be due to the lack of spatial representations of soil texture, soil depth, and topo-
graphic shading in the model. Further model development will be needed to improve this behavior. A logi-
cal next step will be to incorporate a soil evolution model [e.g., Pelletier et al., 2013; Vanwalleghem et al.,
2013] to our existing model.
5. Conclusions
The profound contribution of this study is in advancing the modeling of the ecohydrogeomorphic evolution
of semiarid fluvial catchments with spatial and temporal soil moisture and vegetation dynamics consistent
with generally observed patterns in semiarid ecosystems. For the modeled landscape conditions in central
New Mexico, and by inference for observed landscapes, the main conclusions drawn are:
1. The spatial organization of modeled soil moisture is controlled by: (a) slope and aspect during the drying
stage of the catchment that results in wetter root-zone conditions on NFS as a consequence of their rela-
tively lower solar radiation exposure than SFS; and (b) the hydrologic network connectivity during the
wetting stage of the catchment that leads to rapid wetting patterns in valleys [e.g., Western et al., 2004;
Vivoni et al., 2010]. Throughout the year, aspect-controlled patterns dominate soil moisture dynamics
compared to network-controlled patterns associated with rainfall seasonality.
2. Aspect and network connectivity properties are also manifested on the spatial variation of modeled grass
biomass. NFS exhibit greater biomass productivity than SFS, as they experience longer periods of wetter
conditions than SFS (e.g., Figure 7). Network connectivity, however, showed a markedly different impact
on modeled grass biomass of opposing aspects. While the modeled biomass is greater on steep NFS and
decreases with growing drainage areas in the valleys, modeled SFS biomass is low on hillslopes but con-
sistently grows as a function of drainage areas in the valleys, consistent with empirical investigations
[Flores Cervantes et al., 2014].
3. Coupling ecohydrogeomorphic processes governed by explicit treatment of solar radiation in a landscape
evolution model reveals a positive (negative) feedback mechanism between slope evolution and vegeta-
tion biomass on NFS (SFS), with critical landscape-scale implications. As NFS get steeper by continuing
uplift they support denser vegetation cover due to diminished solar radiation expose with slope, reducing
the efficiency of fluvial erosion and transport, which leads to further slope steepening until erosion and
uplift attains a dynamic equilibrium. Conversely, on SFS, as slopes grow with uplift, increased solar radia-
tion exposure with slope supports sparser biomass that results in more effective fluvial transport, and
resulting in shallower topography when erosion and uplift attains a dynamic equilibrium. At the land-
scape scale, these differential erosion processes lead to asymmetric development of catchment forms,
consistent with regional observations [e.g., Istanbulluoglu et al., 2008].
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1150
While the model provided interesting and promising results, the coupling between geomorphology and
ecohydrology is limited to fluvial processes. This is an important limitation of our model as mounting field
evidences relate colluvial transport and bedrock weathering to vegetation and microclimatic conditions
[Anderson et al., 2013; West et al., 2014]. To incorporate such effects to landscape modeling, Pelletier et al.
[2013] linked colluvial and fluvial transport and bedrock weathering to effective energy and mass transfer
concept which represents the contributions of effective precipitation and biomass production on geomor-
phic processes.
The fact that our model only operates with a single vegetation type remains to be a critical limitation.
In semiarid ecosystems, enhanced sediment transport has been associated with vegetation changes
related to the encroachment of woody plants (shrub encroachment) into grasslands [e.g., Schlesinger
et al., 1990; Van Auken, 2000; Michaelides et al., 2009]. Integration of multiple plant types and the ability
for modeling the competition among tree, grass, and shrub vegetation will open new avenues of
research in regional water, sediment, and nutrient balances in semiarid ecosystems using landscape
models.
Appendix A: Distribution of Solar Radiation on Topography
T
max
on sloping model elements, TS
max , is calculated by scaling the T
max
estimated for a flat surface, TF
max ,
with a solar radiation ratio, R
solar
.R
solar
is defined as the ratio of direct beam solar radiation on sloped surfa-
ces, RS
cs, to that of a flat surface, RF
cs:
TS
max 5TF
max Rsolar 5TF
max RS
cs
RF
cs
;(12a)
The amount of direct beam solar radiation reach at the ground level depends on several factors: the
geometric relations between the Sun and the Earth’s surface, atmospheric attenuation, and topo-
graphic factors [Piedallu and Gegout, 2008]. Ground level direct beam solar radiation, R
gl
, can be esti-
mated as:
Rgl5Isc
dSun
ðÞ
2cos hz
ðÞexp 2n0:12820:054log 10 mðÞðÞm½;(A1)
where I
sc
is the solar constant, 1361 Wm
22
[Kopp and Lean, 2011], d
Sun
is the relative distance between the
Earth and the Sun in astronomical units, cos hz
ðÞis the solar angle of incidence (angle between solar beam
and the normal to the Earth’s surface), mis the optical air mass which is 1=sin !ðÞ,!is solar altitude, nis a
turbidity factor of air (n52 for clear air) [Bras, 1990]. The distance parameter, d
Sun
, can be approximated as:
dSun
ðÞ
25110:033 cos 2p3DOY
365

21
;(A2)
where DOY is day of the year [Duffie and Beckman, 1991].
Topographic factors (hillslope inclination and aspect) are effective on incoming direct beam solar radiation,
R
gl
, by characterizing the solar angle of incidence. The solar angle of incidence is calculated as:
cos hz
ðÞ5cos SðÞsin !ðÞ1sin SðÞcos !ðÞcos w2bðÞ;(A3)
where Sis local slope, !is the solar altitude, wis the azimuth of the Sun, bis aspect, which is an angle
between the direction of the slope face and the geographic North in clockwise rotation. The solar altitude is:
sin !ðÞ5sin KðÞsin dðÞ1cos KðÞcos dðÞcos sh
ðÞ (A4)
where Kis the local latitude, s
h
the hour angle of the Sun, dis the declination angle of the Sun. The azimuth
of Sun is
w5arctan sin sh
ðÞ
tan dðÞcos KðÞ2sin KðÞcos sh
ðÞ

:(A5)
The declination of the Sun, d()is
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1151
d523:45 cos 360
365 1722DOYðÞ

;(A6)
where DOY is the day of year. The hour angle of the Sun, s
h
, is calculated depending on the location of the
Sun. If the sun is east (west) of the local latitude, the hour angle of the Sun is calculated by equation (A7a)
(equation (A7b)
sh5Ts1122DT11DT2
ðÞ315;(A7a)
sh5Ts2122DT11DT2
ðÞ315;(A7b)
where T
s
is the standard time in the time zone of the observer, DT
1
is the time difference between standard
and local magnitude in hours, and DT
2
is the difference between true solar time and mean solar time in
hours, which is usually neglected [Bras, 1990].
The instantaneous direct beam insolation can be integrated for a given finite period of time, when dand K
are constant over this period. Direct beam insolation is estimated each hour of the day, and integrated over
each day of the year. The ratio of RS
cs to RF
cs,R
solar
, is approximately equal to the ratio of incoming solar radia-
tion on inclined surface to the flat surface at noon time RS
gl noon
RF
gl noon

[Flores-Cervantes, 2010]. To reduce the
computational burden in simulations, it is assumed that the ratio of daily solar radiation is equal to the ratio
at noon-time.
Appendix B: Calculation of Potential Transpiration
T
max
is calculated by using the Penman-Monteith (P-M) [Monteith, 1965] transpiration equation for reference
grass [Allen et al., 1989, 1998]. The P-M equation is:
Tmax 5
DRN2GðÞ1qacaes2ea
ðÞ
ra
kvqwD1c11rs
ra
hi
;(B1)
where R
N
is the net radiation at the plant canopy, Gis ground heat flux, c
a
is the specific heat capacity of air, q
a
and q
w
are the density of air and water, respectively, k
v
is the latent heat of vaporization, (e
s
-e
a
) is the vapor
pressure deficit between the leaf and the atmosphere, Dis the slope of the relationship between the saturation
vapor pressure and temperature, cis the psychrometric constant. r
s
and r
a
(s/m) represent plant canopy and aer-
odynamic resistance terms, respectively. r
s
is estimated by scaling up the stomatal resistance of a well-
illuminated leaf, r
l
, with the active (sunlit) live leaf area index, LAI
a,
:r
s
5r
l
/LAI
a
where LAI
a
50.5LAI
Rmax
,r
l
5100
s/m, and LAI
Rmax
52.88 [Allen et al., 1998]. r
a
is calculated based on the von-Karman logarithmic profile as a func-
tion of vegetation height and heights at which wind and relative humidity were measured [Allen et al., 1998].
As expressed in the first term of the P-M equation, T
max
increases with R
N
.R
N
is composed of net short-
wave, R
ns
, and net long-wave radiation, R
nl
:
RN5Rns1Rnl ;(B2)
where, R
ns
is the amount of incoming short-wave radiation received by the surface after a fraction of it is
reflected from the surface as defined by albedo, a(Figure 1):
Rns5ð12aÞRs;(B3)
The incoming short-wave radiation R
s
is the source term in equation (B3). For reference, grass a50.23 [Allen
et al., 1998]. In most land surface models, R
s
is estimated from extraterrestrial radiation, R
ext
, or from the
clear-sky radiation R
cs
.R
ext
is the radiation received at the top of the atmosphere. R
cs
is the fraction of R
ext
retained on the surface after the influence of atmospheric water vapor and dust is taken out. R
s
is the frac-
tion of R
cs
received by the surface reduced by clouds and optical transmission losses. R
s
can be estimated
from extraterrestrial radiation, R
ext
, by relating R
s
to R
ext
through empirical relations that involve the differ-
ence between the minimum and maximum daily temperatures [Hargreaves and Samani, 1982; Thornton and
Running, 1999] or from the clear-sky radiation, R
cs
, scaled with a function for cloud cover [Ivanov et al.,
2004]. Both R
ext
and R
cs
can be estimated as a function of day of year, latitude, local slope, and aspect [Bras,
1990; Dingman, 2002] as described in Appendix A.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1152
The net long-wave radiation, R
nl
, in equation (B2) is the difference between the incoming long-wave radia-
tion from the atmosphere, R
Lin
, and the outgoing long-wave radiation from the Earth surface, R
Lout
(Figure
1). Assuming that the surface temperature is the same as the air temperature, we approximated R
nl
from
the Stefan-Boltzmann law:
Rnl5RLin 2RLout5rT4
Ra2rTa1273:15ðÞ
4;(B4)
where ris the Stefan-Boltzmann constant (5.67 310
28
Wm
22
K
24
), T
a
(K) is the air temperature, and T
Ra
(K)
is the apparent radiative temperature of the atmosphere, which can be calculated by the empirical relation-
ship of [Friend, 1995], TRa5Ta20:825exp 23:54 31022Rs

as a function of shortwave radiation.
For long-term geomorphic evolution simulations driven by generated rainfall, the model is forced by pre-
scribed TF
max , obtained from a cosine function of DOY. The cosine function is fitted to calculated TF
max (by
equation (B.(1)), using observed local climatological data) [Small, 2005]:
TF
max ;DOY 5
Dd
2cos 2pDOY2LT2Nd=2
Nd

1TF
max ;(B5)
where Dd(mm d
21
) is the difference between maximum and minimum values of calibrated daily TF
max
throughout a year, L
T
(d) is the lag between the peak TF
max and peak solar forcing, N
d
is the number of days
in a year, and TF
max is the mean annual of daily TF
max .
References
Allen, R. G., M. E. Jensen, J. L. Wright, and R. D. Burman (1989), Operational estimates of reference evapotranspiration, Agron. J.,81(4), 650–
662.
Allen, R. G., L. S. Pereira, D. Raes, and M. Smith (1998), Crop evapotranspiration: Guidelines for computing crop water requirements, FAO
Irrig. Drain. Pap. 56, Food and Agriculture Organization of the United Nations, Rome, Italy.
Anderson, R. S., S. P. Anderson, and G. E. Tucker (2013), Rock damage and regolith transport by frost: An example of climate modulation of
the geomorphology of the critical zone, Earth Surf. Processes Landforms,38(3), 299–316, doi:10.1002/esp.3330.
Bierman, P. R., J. M. Reuter, K. Pavich, A. C. Gellis, M. W. Caffee, and J. Larsen (2005), Using cosmogenic nuclides to contrast rates of erosion
and sediment yield in a semi-arid, arroyo-dominated landscape, Rio Puerco Basin, New Mexico, Earth Surf. Processes Landforms,30(8),
935–953, doi:10.1002/Esp.1255.
Black, T. A., and D. R. Montgomery (1991), Sediment transport by burrowing mammals, Marin County, California, Earth Surf. Processes Land-
forms,16(2), 163–172, doi:10.1002/esp.3290160207.
Bonta, J. V. (2004), Stochastic simulation of storm occurrence, depth, duration, and within-storm intensities, Trans. ASAE,47(5), 1573–1584.
Bras, R. L. (1990), Hydrology: An Introduction to Hydrologic Science, pp. 19–51, Addison-Wesley, Reading, Mass.
Brolsma, R. J., and M. F. P. Bierkens (2007), Groundwater-soil water-vegetation dynamics in a temperate forest ecosystem along a slope,
Water Resour. Res.,43(1), W01414, doi:10.1029/2005WR004696.
Brutsaert, W. (1982), Evaporation Into the Atmosphere, 299 pp., Reidel Dordrecht, Norwell, Mass.
Burnett, B. N., G. A. Meyer, and L. D. McFadden (2008), Aspect-related microclimatic influences on slope forms and processes, northeastern
Arizona, J. Geophys. Res.,113, F03002, doi:10.1029/2007JF000789.
Butler, J., H. Goetz, and J. L. Richardson (1986), Vegetation and soil: Landscape relationships in the North-Dakota Badlands, Am. Midl. Nat.,
116(2), 378–386.
Cabral, M. C., L. Garrote, R. L. Bras, and D. Entekhabi (1992), A kinematic model of infiltration and runoff generation in layered and sloped
soils, Adv. Water. Resour.,15(5), 311–324, doi:10.1016/0309–1708(92)90017-V.
Cayrol, P., A. Chehbouni, L. Kergoat, G. Dedieu, P. Mordelet, and Y. Nouvellon (2000), Grassland modeling and monitoring with SPOT-4 VEGE-
TATION instrument during the 1997–1999 SALSA experiment, Agric. For. Meteorol.,105(1–3), 91–115, doi:10.1016/S0168–1923(00)00191-X.
Chen, X. Y., Y. Rubin, S. Y. Ma, and D. Baldocchi (2008), Observations and stochastic modeling of soil moisture control on evapotranspira-
tion in a Californian oak savanna, Water Resour. Res.,44, W08409, doi:10.1029/2007WR006646.
Choi, M., and J. M. Jacobs (2007), Soil moisture variability of root zone profiles within SMEX02 remote sensing footprints, Adv. Water
Resour.,30(4), 883–896, doi:10.1016/j.advwatres.2006.07.007.
Clapp, E. M., P. R. Bierman, K. K. Nichols, M. Pavich, and M. Caffee (2001), Rates of sediment supply to arroyos from upland erosion deter-
mined using in situ produced cosmogenic
10
Be and
26
Al, Quat. Res.,55(2), 235–245, doi:10.1006/qres.2000.2211.
Collins, D. B. G., and R. L. Bras (2008), Climatic control of sediment yield in dry lands following climate and land cover change, Water Resour.
Res.,44, W10405, doi:10.1029/2007WR006474.
Collins, D. B. G., and R. L. Bras (2010), Climatic and ecological controls of equilibrium drainage density, relief, and channel concavity in dry
lands, Water Resour. Res.,46, W04508, doi:10.1029/2009WR008615.
Collins, D. B. G., R. L. Bras, and G. E. Tucker (2004), Modeling the effects of vegetation-erosion coupling on landscape evolution, J. Geophys.
Res.,109, F03004, doi:10.1029/2003JF000028.
Connel S. D., and D. J. McCraw (2007), Preliminary Geologic Map of the La Joya NW Quadrangle, Socorro County, New Mexico, N. M. Bur. of
Geol. and Miner. Resour., Socorro.
Coppus, R., and A. C. Imeson (2002), Extreme events controlling erosion and sediment transport in a semi-arid sub-Andean valley, Earth
Surf. Processes Landforms,27(13), 1365–1375, doi:10.1002/Esp.435.
Dethier, D. P. (2001), Pleistocene incision rates in the western United States calibrated using Lava Creek B tephra, Geology,29(9), 783–786,
doi:10.1130/0091–7613(2001)029<0783:PIRITW>2.0.CO;2.
Dietrich, W. E., and J. T. Perron (2006), The search for a topographic signature of life, Nature,439, 411–418.
Acknowledgments
The data cited in this manuscript are
publicly available and can be accessed
from the following repositories:
(1) Meteorology data, soil moisture
data, and Bowen Ratio
evapotranspiration data for Deep Well
meteorological station are available at
Sevilleta Long Term Ecological
Research database; their
corresponding data set names are
Sev001, Sev078, and Sev079,
respectively. (2) MODIS LAI data are
available at NASA LP DAAC (Land
Processes Distributed Archive Center)
data repository (https://lpdaac.usgs.
gov/data_access); data set name:
MOD15A2. (3) Walnut Gulch runoff
data are available at Southwest
Watershed Research Center’s online
data repository (http://www.tucson.ars.
ag.gov/dap/). We thank J. Bruce J.
Harrison, Hugo A. Guti
errez-Jurado,
Xiaochi Zhou, Valeriy Y. Ivanov, Tiejun
Wang, and Ronda L. Strauch for their
collaboration and constructive
comments. We also thank the
Associate Editor P. Saco and three
anonymous reviewers for their
comments, which contributed to
improving this paper. We thank NSF
and NASA for financial support
provided through grants: NSF-EAR
0963858, NSF-ACI 1148305, NSF-EAR
0819924, and NASA NNG05GA17G.
Yetemen acknowledges support from
the GSA Farouk El-Baz student
research award. Prof. Bras thanks the K.
Harrison Brown Chair at Georgia Tech
that facilitates his research.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1153
Dietrich, W. E., D. G. Bellugi, L. S. Sklar, J. D. Stock, A. M. Heimsath, and J. J. Roering (2003), Geomorphic transport laws for predicting land-
scape form and dynamics, in Prediction in Geomorphology, Geophys. Monogr. Ser., vol. 135, edited by P. R. Wilcock and R. M. Iverson,
pp. 103–132, AGU, Washington, D. C.
Dingman, S. L. (2002), Physical Hydrology, 2nd ed., pp. 601–607, Prentice Hall, Upper Saddle River, N. J.
Drissel, J. C., and H. B. Osborn (1968), Variability in rainfall producing runoff from a semiarid rangeland watershed, Alamogordo Creek, New
Mexico, J. Hydrol.,6(2), 194–201, doi:10.1016/0022-1694(68)90159-5.
Duffie, J. A., and W. A. Beckman (1991), Solar Engineering of Thermal Processes, 2nd ed., 10 pp., John Wiley, N. Y.
Dunkerley, D. (2010), Ecogeomorphology in the Australian drylands and the role of biota in mediating the effects of climate change on
landscape processes and evolution, Geol. Soc. London Spec. Publ.,346(1), 87–120, doi:10.1144/sp346.7.
Dunkerley, D. L. (1992), Channel geometry, bed material, and inferred flow conditions in ephemeral stream systems, Barrier Range, Western
NSW Australia, Hydrol. Processes,6(4), 417–433, doi:10.1002/hyp.3360060404.
Dunne, T., W. H. Zhang, and B. F. Aubry (1991), Effects of rainfall, vegetation, and microtopography on infiltration and runoff, Water Resour.
Res.,27(9), 2271–2285, doi:10.1029/91WR01585.
Eagleson, P. S. (1978), Climate, soil, and vegetation 2. The distribution of annual precipitation derived from observed storm sequences,
Water Resour. Res.,14(5), 713–721, doi:10.1029/WR014i005p00713.
Einstein, H. A., and N. L. Barbarossa (1952), River channel roughness, Trans. Am. Soc. Civ. Eng.,117, 1121–1132.
Engman, E. T. (1986), Roughness coefficients for routing surface runoff, J. Irrig. Drain. Eng.,112(1), 39–53.
Famiglietti, J. S., D. R. Ryu, A. A. Berg, M. Rodell, and T. J. Jackson (2008), Field observations of soil moisture variability across scales, Water
Resour. Res.,44, W01423, doi:10.1029/2006WR005804.
Fang, H. L., and S. L. Liang (2005), A hybrid inversion method for mapping leaf area index from MODIS data: Experiments and application
to broadleaf and needle leaf canopies, Remote Sens. Environ.,94(3), 405–424, doi:10.1016/j.rse.2004.11.001.
Flores-Cervantes, J. H. (2010), The coupled development of terrain and vegetation: The case of semiarid grasslands, PhD thesis, MIT, Bos-
ton, Mass.
Flores Cervantes, J. H., E. Istanbulluoglu, E. R. Vivoni, C. D. Holifield Collins, and R. L. Bras (2014), A geomorphic perspective on terrain-
modulated organization of vegetation productivity: Analysis in two semiarid grassland ecosystems in Southwestern United States, Eco-
hydrology,7(2), 242–257, doi:10.1002/eco.1333.
Foster, G. R. (1982), Modeling the erosion process, in Hydrologic Modeling of Small Watersheds, ASAE Monogr., vol. 5, edited by C. T. Haan,
pp. 295–380, Am. Soc. Agric. Eng., St. Joseph, Miss.
Friend, A. D. (1995), PGEN: an integrated model of leaf photosynthesis, transpiration, and conductance, Ecol. Modell.,77(2–3), 233–255, doi:
10.1016/0304-3800(93)E0082-E.
Gabet, E. J. (2003), Sediment transport by dry ravel, J. Geophys. Res.,108(B1), 2049, doi:10.1029/2001JB001686.
Gasparini, N. M., R. L. Bras, and K. X. Whipple (2006), Numerical modeling of non-steady-state river profile evolution using a sediment-flux-
dependent incision model, in Tectonics, Climate and Landscape Evolution, edited by S. D. Willett et al., Spec. Pap. Geol. Soc. Am., 398,
127–141.
Goodrich, D. C., T. O. Keefer, C. L. Unkrich, M. H. Nichols, H. B. Osborn, J. J. Stone, and J. R. Smith (2008), Long-term precipitation database,
Walnut Gulch Experimental Watershed, Arizona, United States, Water Resour. Res.,44, W05S04, doi:10.1029/2006WR005782.
Gosz, J. R. (2012a), Bowen Ratio evapotranspiration data at the Sevilleta National Wildlife Refuge, New Mexico, 1996–1999, Sevilleta Long
Term Ecological Research Database: SEV079, Albuquerque, N.M. [Available at http://sev.lternet .edu/data/sev-79, last accessed 12 Jan
2014.]
Gosz, J. R. (2012b), Time Domain Reflectometry at the Sevilleta National Wildlife Refuge, New Mexico, 1996–2005, Sevilleta Long Term Ecologi-
cal Research Database: SEV078, Albuquerque, N.M. [Available at http://sev.lternet.edu/data/sev-78, last accessed 12 Jan 2014.]
Gosz, J. R., D. I. Moore, G. A. Shore, H. D. Grover, W. Rison, and C. Rison (1995), Lightning estimates of precipitation location and quantity on
the Sevilleta LTER, New Mexico, Ecol. Appl.,5(4), 1141–1150, doi:10.2307/2269361.
Govers, G. (1992), Evaluation of transporting capacity formulae for overland flow conditions, in Overland Flow, Hydraulics and Erosion
Mechanics, edited by A. J. Parsons and A. D. Abrahams, pp. 243–273, UCL Press, London, U. K.
Grayman, W. M., and P. S. Eagleson (1969), Streamflow Record Length for Modelling Catchment Dynamics, MIT Press, Cambridge, Mass.
Gregory, K. J. (1976), Drainage networks and climate, in Geomorphology and Climate, edited by E. Derbyshire, pp. 289–315, John Wiley, Lon-
don, U. K.
Guti
errez-Jurado, H. A., and E. R. Vivoni (2013), Ecogeomorphic expressions of an aspect-controlled semiarid basin: I. Topographic analyses
with high-resolution data sets, Ecohydrology,6(1), 8–23, doi:10.1002/Eco.280.
Guti
errez-Jurado, H. A., E. R. Vivoni, J. B. J. Harrison, and H. Guan (2006), Ecohydrology of root zone water fluxes and soil development in
complex semiarid rangelands, Hydrol. Processes,20(15), 3289–3316, doi:10.1002/Hyp.6333.
Guti
errez-Jurado, H. A., E. R. Vivoni, E. Istanbulluoglu, and R. L. Bras (2007), Ecohydrological response to a geomorphically significant flood
event in a semiarid catchment with contrasting ecosystems, Geophys. Res. Lett.,34, L24S25, doi:10.1029/2007GL030994.
Guti
errez-Jurado, H. A., E. R. Vivoni, C. Cikoski, J. B. J. Harrison, R. L. Bras, and E. Istanbulluoglu (2013), On the observed ecohydrologic
dynamics of a semiarid basin with aspect-delimited ecosystems, Water Resour. Res.,49, 8263–8284, doi:10.1002/2013WR014364.
Hanks, T. C. (2000), The age of scarp-like landforms from diffusion-equation analysis, in Quaternary Geochronology: Methods and Applica-
tions, vol. 4, edited by J. S. Noller, J. M. Sowers, and W. R. Lettis, pp. 313–338, AGU, Washington, D. C.
Hargreaves, G. H., and Z. A. Samani (1982), Estimating potential evapotranspiration, J. Irrig. Drain. Div.,108(3), 225–230.
Hillel, D. (2006), The Natural History of the Bible: An EnvironmentalExploration of the Hebrew Scriptures, pp. 149–150, Columbia Univ. Press, N. Y.
Hughes, M., P. Almond, and J. J. Roering (2009), Increased sediment transport via bioturbation at the last glacial-interglacial transition, Geol-
ogy,37, 919–922, doi:10.1130/G30159A.1.
Hurst, M. D., S. M. Mudd, K. Yoo, M. Attal, and R. Walcott (2013), Influence of lithology on hillslope morphology and response to tectonic
forcing in the northern Sierra Nevada of California, J. Geophys. Res. Earth Surf.,118, 832–851, doi:10.1002/jgrf.20049.
Istanbulluoglu, E., and R. L. Bras (2005), Vegetation-modulated landscape evolution: Effects of vegetation on landscape processes, drainage
density, and topography, J. Geophys. Res.,110, F02012, doi:10.1029/2004JF000249.
Istanbulluoglu, E., and R. L. Bras (2006), On the dynamics of soil moisture, vegetation, and erosion: Implications of climate variability and
change, Water Resour. Res.,42, W06418, doi:10.1029/2005WR004113.
Istanbulluoglu, E., D. G. Tarboton, R. T. Pack, and C. Luce (2003), A sediment transport model for incision of gullies on steep topography,
Water Resour. Res.,39(4), 1103, doi:10.1029/2002WR001467.
Istanbulluoglu, E., O. Yetemen, E. R. Vivoni, H. A. Gutierrez-Jurado, and R. L. Bras (2008), Eco-geomorphic implications of hillslope aspect:
Inferences from analysis of landscape morphology in central New Mexico, Geophys. Res. Lett.,35, L14403, doi:10.1029/2008GL034477.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1154
Istanbulluoglu, E., T. J. Wang, and D. A. Wedin (2012), Evaluation of ecohydrologic model parsimony at local and regional scales in a semi-
arid grassland ecosystem, Ecohydrology,5(1), 121–142, doi:10.1002/Eco.211.
Ivanov, V. Y., E. R. Vivoni, R. L. Bras, and D. Entekhabi (2004), Catchment hydrologic response with a fully distributed triangulated irregular
network model, Water Resour. Res.,40, W11102, doi:10.1029/2004WR003218.
Ivanov, V. Y., R. L. Bras, and D. C. Curtis (2007), A weather generator for hydrological, ecological, and agricultural applications, Water Resour.
Res.,43, W10406, doi:10.1029/2006WR005364.
Ivanov, V. Y., R. L. Bras, and E. R. Vivoni (2008a), Vegetation-hydrology dynamics in complex terrain of semiarid areas: 1. A mechanistic
approach to modeling dynamic feedbacks, Water Resour. Res.,44, W03429, doi:10.1029/2006WR005588.
Ivanov, V. Y., R. L. Bras, and E. R. Vivoni (2008b), Vegetation-hydrology dynamics in complex terrain of semiarid areas: 2. Energy-water con-
trols of vegetation spatiotemporal dynamics and topographic niches of favorability, Water Resour. Res.,44, W03430, doi:10.1029/
2006WR005595.
Ivanov, V. Y., S. Fatichi, G. D. Jenerette, J. F. Espeleta, P. A. Troch, and T. E. Huxman (2010), Hysteresis of soil moisture spatial heterogeneity
and the "homogenizing" effect of vegetation, Water Resour. Res.,46, W09521, doi:10.1029/2009WR008611.
Kirkby, M. J. (1971), Hillslope process-response models based on the continuity equation, Spec. Publ. Inst. Br. Geogr.,3, 15–30.
Kopp, G., and J. L. Lean (2011), A new, lower value of total solar irradiance: Evidence and climate significance, Geophys. Res. Lett.,38,
L01706, doi:10.1029/2010GL045777.
Kunkel, M. L., A. N. Flores, T. J. Smith, J. P. McNamara, and S. G. Benner (2011), A simplified approach for estimating soil carbon and nitrogen
stocks in semi-arid complex terrain, Geoderma,165(1), 1–11, doi:10.1016/j.geoderma.2011.06.011.
Kurc, S. A., and E. E. Small (2004), Dynamics of evapotranspiration in semiarid grassland and shrubland ecosystems during the summer
monsoon season, central New Mexico, Water Resour. Res.,40, W09305, doi:10.1029/2004WR003068.
Laio, F., A. Porporato, L. Ridolfi, and I. Rodriguez-Iturbe (2001), Plants in water-controlled ecosystems: Active role in hydrologic processes
and response to water stress II. Probabilistic soil moisture dynamics, Adv. Water Resour.,24(7), 707–723, doi:10.1016/S0309-
1708(01)00005–7.
Larcher, W. (1995), Physiological Plant Ecology: Ecophysiology and Stress Physiology of Functional Groups, 3rd ed., 122 pp., Springer, N. Y.
Langbein, W. B., and S. Schumm (1958), Yield of sediment in relation to mean annual precipitation, Trans. AGU,39, 1076–1084.
Lauenroth, W. K., and J. B. Bradford (2009), Ecohydrology of dry regions of the United States: Precipitation pulses and intraseasonal
drought, Ecohydrology,2(2), 173–181, doi:10.1002/Eco.53.
Laursen, E. M. (1958), The total sediment load of streams, J. Hydraul. Div. Am. Soc. Civ. Eng.,84 (1530), 1–6.
Lawrence, J. E., and G. M. Hornberger (2007), Soil moisture variability across climate zones, Geophys. Res. Lett.,34, L20402, doi:10.1029/
2007GL031382.
Lee, T. J. (1992), The impact of vegetation on the atmospheric boundary layer and convective storms, PhD dissertation thesis, Colo. State
Univ., Fort Collins, Colo.
Levins, R. (1969), Some demographic and genetic consequences of environmental heterogeneity for biological control, Bull. Entomol. Soc.
Am.,15(3), 237–240.
Loik, M. E., D. D. Breshears, W. K. Lauenroth, and J. Belnap (2004), A multi-scale perspective of water pulses in dryland ecosystems: Climatol-
ogy and ecohydrology of the western USA, Oecologia,141(2), 269–281, doi:10.1007/s00442-004-1570-y.
Ma, L., L. X. Jin, and S. L. Brantley (2011), How mineralogy and slope aspect affect REE release and fractionation during shale weathering in
the Susquehanna/Shale Hills Critical Zone Observatory, Chem. Geol.,290(1-2), 31–49, doi:10.1016/j.chemgeo.2011.08.013.
Mascaro, G., E. R. Vivoni, and R. Deidda (2011), Soil moisture downscaling across climate regions and its emergent properties, J. Geophys.
Res.,116, D22114, doi:10.1029/2011JD016231.
McGuire, L. A., J. D. Pelletier, and J. J Roering (2014), Development of topographic asymmetry: Insights from dated cinder cones in the
western United States, J. Geophys. Res. Earth Surf.,119(8), 1725–1750, doi:10.1002/2014JF003081.
McMahon, D. R. (1998), Soil, Landscape and Vegetation Interactions in Small Semi-Arid Drainage Basin: Sevilleta National Wildlife Refuge, New
Mexico, NMTech, Socorro, N. M.
Meyer-Peter, E., and R. M
uller (1948), Formulas for bed load transport, in Proceeding of the 2nd Meeting of the International Association for
Hydraulic Structures Research, pp. 39–64, International Association for Hydraulic Structures Research, Stockholm, Sweden.
Michaelides, K., D. Lister, J. Wainwright, and A. J. Parsons (2009), Vegetation controls on small-scale runoff and erosion dynamics in a
degrading dryland environment, Hydrol. Processes,23, 1617–1630, doi:10.1002/hyp.7293.
Milne, B. T., D. I. Moore, J. L. Betancourt, J. A. Fairchild-Parks, T. W. Swetnam, R. R. Parmenter, and W. T. Pockman (2003), Multidecadal
drought cycles in South-central New Mexico: Patterns and consequences, in Climate Variabililty and Ecosystem Response at Long Term
Ecological Research (LTER) Sites, edited by D. Greenland, D. Goodin, and R. Smith, pp. 286–307, Oxford Univ. Press, N. Y.
Moglen, G. E., E. A. B. Eltahir, and R. L. Bras (1998), On the sensitivity of drainage density to climate change, Water Resour. Res.,34(4),
855–862, doi:10.1029/97WR02709.
Moln
ar, P., and J. A. Ram
ırez (2001), Recent trends in precipitation and streamflow in the Rio Puerco Basin, J Clim.,14(10), 2317–2328, doi:
10.1175/1520-0442(2001)014<2317:RTIPAS>2.0.CO;2.
Monson, R. K., M. R. Sackschewsky, and G. J. Williams III (1986), Field measurements of photosynthesis, water-use efficiency, and growth in
Agropyron smithii (C
3
) and Bouteloua gracilis (C
4
) in the Colorado shortgrass steppe, Oecologia,68, 400–409, doi:10.1007/BF01036746.
Montaldo, N., R. Rondena, J. D. Albertson, and M. Mancini (2005), Parsimonious modeling of vegetation dynamics for ecohydrologic studies
of water-limited ecosystems, Water Resour. Res.,41, W10416, doi:10.1029/2005WR004094.
Monteith, J. L. (1965), Evaporation and environment, Symp. Soc. Exp. Biol.,19, 205–234.
Montgomery, D. R (2001), Slope distributions, threshold hillslopes and steady-state topography, Am. J. Sci.,301, 432–454, doi:10.2475/
ajs.301.4-5.432.
Moore, D. I. (2012), Meteorology data at the Sevilleta National Wildlife Refuge,New Mexico, Sevilleta Long Term Ecological Research Data-
base: SEV001, Albuquerque, N. M. [Available at http://sev.lternet.edu/data/sev-1, last accessed 12 Jan 2014.]
Mudd, S. M. (2006), Investigation of the hydrodynamics of flash floods in ephemeral channels: Scaling analysis and simulation using a shock-
capturing flow model incorporating the effects of transmission losses, J. Hydrol.,324(1-4), 65–79, doi:10.1016/j.jhydrol.2005.09.012.
Mutziger, A. J., C. M. Burt, D. J. Howes, and R. G. Allen (2005), Comparison of measured and FAO-56 modeled evaporation from bare soil, J.
Irrig. Drain. Eng.,131(1), 59–72, doi:10.1061/(ASCE)0733-9437(2005)131:1(59).
Nash, D. B. (1980), Forms of bluffs degraded for different lengths of time in Emmett County, Michigan, USA, Earth Surf. Processes Landforms,
5, 331–345.
Nearing, M. A., J. R. Simanton, L. D. Norton, S. J. Bulygin, and J. Stone (1999), Soil erosion by surface water flow on a stony, semiarid hill-
slope, Earth Surf. Processes Landforms,24(8), 677–686, doi:10.1002/(SICI)1096-9837(199908)24:8<677::AID-ESP981>3.3.CO;2-T.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1155
Parsons, A. J., J. Wainwright, P. M. Stone, and A. D. Abrahams (1999), Transmission losses in rills on dryland hillslopes, Hydrol. Processes,
13(17), 2897–2905, doi:10.1002/(SICI)1099-1085(19991215)13:17<2897::AID-HYP905>3.0.CO;2-B.
Pelletier, J. D., et al. (2013), Coevolution of nonlinear trends in vegetation, soils, and topography with elevation and slope aspect: A case
study in the sky islands of southern Arizona, J. Geophys. Res. Earth Surf.,118, 741–758, doi:10.1002/Jgrf.20046.
Penna, D., H. J. Tromp-van Meerveld, A. Gobbi, M. Borga, and G. Dalla Fontana (2011), The influence of soil moisture on threshold runoff
generation processes in an alpine headwater catchment, Hydrol. Earth Syst. Sci.,15(3), 689–702, doi:10.5194/hess-15-689-2011.
Perron, J. T., W. E. Dietrich, and J. W. Kirchner (2008), Controls on the spacing of first-order valleys, J. Geophys. Res.,113, F04016, doi:
10.1029/2007JF000977.
Piedallu, C., and J. Gegout (2008), Efficient assessment of topographic solar radiation to improve plant distribution models, Agric. For. Mete-
orol.,148(11), 1696–1706, doi:0.1016/j.agrformet.2008.06.001.
Pierce, J. K. B., T. Lookingbill, and D. Urban (2005), A simple method for estimating potential relative radiation (PRR) for landscape-scale
vegetation analysis, Landscape Ecol.,20(2), 137–147, doi:10.1007/s10980-004-1296-6.
Pierce, K. L., and S. M. Colman (1986), Effect of height and orientation (microclimate) on geomorphic degradation rates and processes,
late-glacial terrace scarps in central Idaho, Geol. Soc. Am. Bull.,97(7), 869–885, doi:10.1130/0016-7606(1986)97 <869:EOHAOM>2.0.CO;2.
Polyakov, V. O., M. A. Nearing, M. H. Nichols, R. L. Scott, J. J. Stone, and M. P. McClaran (2010), Long-term runoff and sediment yields from
small semiarid watersheds in southern Arizona, Water Resour. Res.,46, W09512, doi:10.1029/2009WR009001.
Porporato, A., E. Daly, and I. Rodriguez-Iturbe (2004), Soil water balance and ecosystem response to climate change, Am. Nat.,164(5),
625–632, doi:10.1086/424970.
Poulos, M. J., J. L. Pierce, A. N. Flores, and S. G. Benner (2012), Hillslope asymmetry maps reveal widespread, multi-scale organization, Geo-
phys. Res. Lett.,39, L06406, doi:10.1029/2012GL051283.
Roering, J. J., J. W. Kirchner, and W. E. Dietrich (1999), Evidence for nonlinear, diffusive sediment transport on hillslopes and implications
for landscape morphology, Water Resour. Res.,35(3), 853–870, doi:10.1029/1998WR900090.
Roering, J. J., J. T. Perron, and J. W. Kirchner (2007), Functional relationships between denudation and hillslope form and relief, Earth Planet.
Sci. Lett.,264(1-2), 245–258, doi:10.1016/j.epsl.2007.09.035.
Rosenbaum, U., H. R. Bogena, M. Herbst, J. A. Huisman, T. J. Peterson, A. Weuthen,A. W. Western, and H. Vereecken (2012), Seasonal and event
dynamics of spatial soil moisture patterns at the small catchment scale, Water Resour. Res.,48, W10544, doi:10.1029/2011WR011518.
Ryu, D., and J. S. Famiglietti (2005), Characterization of footprint-scale surface soil moisture variability using Gaussian and beta distribution
functions during the Southern Great Plains 1997 (SGP97) hydrology experiment, Water Resour. Res.,41, W12433, doi:10.1029/
2004WR003835.
Saco, P. M., G. R. Willgoose, and G. R. Hancock (2007), Eco-geomorphology of banded vegetation patterns in arid and semi-arid regions,
Hydrol. Earth Syst. Sci.,11(6), 1717–1730, doi:10.5194/hess-11-1717-2007.
Sala, O. E., and W. K. Lauenroth (1982), Small rainfall events: An ecological role in semiarid regions, Oecologia,53(3), 301–304, doi:10.1007/
Bf00389004.
Schlesinger, W. H., J. F. Reynolds, G. L. Cunningham, L. F. Huenneke, W. M. Jarrell, R. A. Virginia, and W. G. Whitford (1990), Biological feed-
backs in global desertification, Science,247(4946), 1043–1048, doi:10.1126/science.247.4946. 1043.
Scholes, R. J., and B. H. Walker (1993), An African Savanna: Synthesis of the Nylsvley Study, 306 pp., Cambridge Univ. Press, Cambridge, U. K.
Shabanov, N. V., et al. (2005), Analysis and optimization of the MODIS leaf area index algorithm retrievals over broadleaf forests, IEEE Trans.
Geosci. Remote Sens.,43(8), 1855–1865, doi:10.1109/Tgrs.2005.852477.
Shields, A. (1936), Application of similarity principles and turbulence research to bed-load movement [in German], No. 26, Preuss. Vers.
Anst. Wasserb. Schiffb. [English translation is available as Hydrodyn. Lab. Publ. 167, Hydrodyn. Lab., Calif. Inst. of Technol., Pasadena].
Simons, D. B., and F. S¸ent
urk (1992), Sediment Transport Technology: Water and Sediment Dynamics, Water Resour. Publ., Littleton, Colo.
Sitch, S., et al. (2003), Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegeta-
tion model, Global Change Biol.,9(2), 161–185, doi:10.1046/j.1365-2486.2003.00569.x.
Small, E. E. (2005), Climatic controls on diffuse groundwater recharge in semiarid environments of the southw estern United States, Water
Resour. Res.,41, W04012, doi:10.1029/2004WR003193.
Soil Survey Staff (2012), Natural Resources Conservation Service, Web Soil Survey, U. S. Dep. of Agric. [Available at http://websoilsurvey.
nrcs.usda.gov/, Last accessed 27 May 2012.]
Stone, J. J., M. H. Nichols, D. C. Goodrich, and J. Buono (2008), Long-term runoff database, Walnut Gulch Experimental Watershed, Arizona,
United States, Water Resour. Res.,44, W05S05, doi:10.1029/2006WR005733.
Summerfield, M. A., and N. J. Hulton (1994), Natural controls of fluvial denudation rates in major world drainage basins, J. Geophys. Res.,
99(B7), 13,871–13,883, doi:10.1029/94JB00715.
Swenson, J. J., and R. H. Waring (2006), Modelled photosynthesis predicts woody plant richness at three geographic scales across the
north-western United States, Global Ecol. Biogeogr.,15(5), 470–485, doi:10.1111/j.1466-822x.2006.00242.x.
Tarboton, D. G., R. L. Bras, and I. Rodriguez-Iturbe (1991), On the extraction of channel networks from digital elevation data, Hydrol. Proc-
esses,5(1), 81–100, doi:10.1002/hyp.3360050107.
Teuling, A. J., F. Hupet, R. Uijlenhoet, and P. A. Troch (2007), Climate variability effects on spatial soil moisture dynamics, Geophys. Res. Lett.,
34, L06406, 10.1029/2006GL029080.
Thornes, J. B. (1990), The interaction of erosional and vegetational dynamics in land degradation: Spatial outcomes, in Vegetation and Ero-
sion, edited by J. B. Thornes, pp. 41–53, John Wiley, Chichester, U. K.
Thornton, P. E., and S. W. Running (1999), An improved algorithm for estimating incident daily solar radiation from measurements of tem-
perature, humidity, and precipitation, Agric. For. Meteorol.,93(4), 211–228, doi:10.1016/S0168-1923(98)00126-9.
Tucker, G. E., and R. L. Bras (1998), Hillslope processes, drainage density, and landscape morphology, Water Resour. Res.,34(10), 2751–2764,
doi:10.1029/98WR01474.
Tucker, G. E., and R. L. Bras (2000), A stochastic approach to modeling the role of rainfall variability in drainage basin evolution, Water
Resour. Res.,36(7), 1953–1964, doi:10.1029/2000WR900065.
Tucker, G. E., and G. R. Hancock (2010), Modelling landscape evolution, Earth Surf. Processes Landforms,35(1), 28–50, doi:10.1002/Esp.1952.
Tucker, G. E., and K. X. Whipple (2002), Topographic outcomes predicted by stream erosion models: Sensitivity analysis and intermodel
comparison, J. Geophys. Res.,107(B9), 2179, doi:10.1029/2001JB000162.
Tucker, G. E., S. T. Lancaster, N. M. Gasparini, and R. L. Bras (2001a), The Channel-Hillslope Integrated Landscape Development model
(CHILD), in Landscape Erosion and Evolution Modeling, edited by R. S. Harmon and W. W. Doe III, pp. 349–388, Kluwer Acad., N. Y.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1156
Tucker, G. E., S. T. Lancaster, N. M. Gasparini, R. L. Bras, and S. M. Rybarczyk (2001b), An object-oriented framework for distributed hydro-
logic and geomorphic modeling using triangulated irregular networks, Comput. Geosci.,27(8), 959–973, doi:10.1016/S0098-
3004(00)00134-5.
Tucker, G. E., L. Arnold, R. L. Bras, H. Flores, E Istanbulluoglu, and P. S
olyom (2006), Headwater channel dynamics in semiarid rangelands,
Colorado high plains, USA, Geol. Soc. Am. Bull.,118, 959–974, doi:10.1130/B25928.25921.
Vanwalleghem, T., U. Stockmann, B. Minasny, and A. B. McBratney (2013), A quantitative model for integrating landscape evolution and
soil formation, J. Geophys. Res. Earth Surf.,118, 331–347, doi:10.1029/2011jf002296.
Van Auken, O. W. (2000), Shrub invasions of North American semiarid grasslands, Annu. Rev. Ecol. Syst.,31(1), 197–215, doi:10.1146/
annurev.ecolsys.31.1.197.
Vereecken, H., T. Kamai, T. Harter, R. Kasteel, J. Hopmans, and J. Vanderborght (2007), Explaining soil moisture variability as a function of
mean soil moisture: A stochastic unsaturated flow perspective, Geophys. Res. Lett.,34, L22402, doi:10.1029/2007GL03181 3.
Vivoni, E. R., R. S. Bowman, R. L. Wyckoff, R. T. Jakubowski, and K. E. Richards (2006), Analysis of a monsoon flood event in an ephemeral
tributary and its downstream hydrologic effects, Water Resour. Res.,42, W03404, doi:10.1029/2005WR004036.
Vivoni, E. R., H. A. Moreno, G. Mascaro, J. C. Rodriguez, C. J. Watts, J. Garatuza-Payan, and R. L. Scott (2008), Observed relation between evapo-
transpiration and soil moisture in the North American monsoon region, Geophys. Res. Lett.,35, L22403, doi:10.1029/2008GL036001.
Vivoni, E. R., K. Tai, and D. J. Gochis (2009), Effects of initial soil moisture on rainfall generation and subsequent hydrologic response during
the North American Monsoon, J. Hydrometeorol.,10(3), 644–664, doi:10.1175/2008jhm1069.1.
Vivoni, E. R., J. C. Rodriguez, and C. J. Watts (2010), On the spatiotemporal variability of soil moisture and evapotranspiration in a mountain-
ous basin within the North American monsoon region, Water Resour. Res.,46, W02509, doi:10.1029/2009WR008240.
Wainwright, J., and A. J. Parsons (2002), The effect of temporal variations in rainfall on scale dependency in runoff coefficients, Water
Resour. Res.,38(12), 1271, doi:10.1029/2000WR000188.
West, N., E. Kirby, P. Bierman, and B. A. Clarke (2014), Aspect-dependent variations in regolith creep revealed by meteoric
10
Be, Geology,42,
507–510, doi:10.1130/g35357.1.
Western, A. W., R. B. Grayson, G. Bl
oschl, and D. J. Wilson (2003), Spatial variability of soil moisture and its implications for scaling, in Scaling
Methods in Soil Physics, edited by Y. Perchepsky et al., pp. 119–142, CRC Press, Boca Raton, Fla.
Western, A. W., S. L. Zhou, R. B. Grayson, T. A. McMahon, G. Bloschl, and D. J. Wilson (2004). Spatial correlation of soil moisture in small
catchments and its relationship to dominant spatial hydrological processes, J. Hydrol.,286, 113–134.
Wigmosta, M. S., L. W. Vail, and D. P. Lettenmaier (1994), A distributed hydrology-vegetation model for complex terrain, Water Resour. Res.,
30(6), 1665–1679, 10.1029/94WR00436.
Willgoose, G. R. (1994), A physical explanation for an observed area-slope-elevation relationship for catchments with declining relief, Water
Resour. Res.,30, 151–159, doi:10.1029/93WR01810.
Willgoose, G., R. L. Bras, and I. Rodriguez-Iturbe (1991), A coupled channel network growth and hillslope evolution model .1. Theory, Water
Resour. Res.,27(7), 1671–1684, doi:10.1029/91WR00935.
Wilson, L. (1973), Variations in mean annual sediment yield as a function of mean annual precipitation, Am. J. Sci.,273(4), 335–349.
Yetemen, O. (2014), Modeling the ecohydrologic role of solar radiation on catchment development in semiarid ecosystems, PhD thesis,
Univ. of Wash., Seattle.
Yetemen, O., E. Istanbulluoglu, and E. R. Vivoni (2010), The implications of geology, soils, and vegetation on landscape morphology: Infer-
ences from semi-arid basins with complex vegetation patterns in Central New Mexico, USA, Geomorpholo gy,116(3-4), 246–263, doi:
10.1016/j.geomorph.2009.11.026.
Zhou, X., E. Istanbulluoglu, and E. R. Vivoni (2013), Modeling the ecohydrological role of aspect-controlled radiation on tree-grass-shrub
coexistence in a semiarid climate, Water Resour. Res.,49(5), 2872–2895, doi:10.1002/wrcr.20259.
Zou, C. B., G. A. Barron-Gafford, and D. D. Breshears (2007), Effects of topography and woody plant canopy cover on near-ground solar radi-
ation: Relevant energy inputs for ecohydrology and hydropedology, Geophys. Res. Lett.,34(24), L24S21, doi:10.1029/2007GL031484.
Water Resources Research 10.1002/2014WR016169
YETEMEN ET AL. V
C2015. American Geophysical Union. All Rights Reserved. 1157
... Vegetation plays a key role in landscape evolution, as it modulates geomorphic processes, such as erosion and sediment transport (Collins & Bras, 2008;Dietrich & Perron, 2006;Kirkby, 1995;Langbein & Schumm, 1958;Moglen & Parsons, 1998;Saco & Moreno-de las Heras, 2013;Willgoose, 2018;Yetemen, Istanbulluoglu, Flores-Cervantes, et al., 2015). Early work by Langbein and Schumm (1958) quantified the non-linear relationship between precipitation and sediment yield and found that vegetation and precipitation exert competing effects, as precipitation increases and vegetation inhibits erosion. ...
... Further work incorporating the effect of vegetation into landscape evolution models (LEMs) has revealed an increase in dynamic equilibrium slopes with increased vegetation cover, which is due to an increase in resistance to sediment transport and erosion (see e.g. Collins & Bras, 2008;Istanbulluoglu & Bras, 2005;Saco et al., 2007;Yetemen, Istanbulluoglu, Flores-Cervantes, et al., 2015;Yetemen, Saco, & Istanbulluoglu, 2019). ...
... Several modelling studies (Baartman et al., 2018;Collins & Bras, 2008Collins et al., 2004;Istanbulluoglu & Bras, 2005;Saco & Moreno-de las Heras, 2013;Saco et al., 2007;Yetemen, Istanbulluoglu, & Duvall, 2015;Yetemen, Istanbulluoglu, Flores-Cervantes, et al., 2015) have investigated the coevolution of vegetation and landforms under different climatic and anthropic conditions. Collins et al. (2004) and Istanbulluoglu and Bras (2005) coupled the Channel-Hillslope Integrated Landscape Development (CHILD) LEM with vegetation-erosion dynamics and found that the dynamic vegetation cover (in which rainfall and solar radiation drive vegetation growth and senescence) led to the formation of a highly dissected topography with a significantly lower relief than the landscape evolved under static vegetation (vegetation remains constant). ...
Article
Topography affects the intensity and spatial distribution of precipitation due to orographic lifting mechanisms and, in turn, influences the prevailing climate and vegetation distribution. Previous modelling studies on the impact of orographic precipitation on landform evolution have considered bare‐soil conditions. However, research on the effect of changes in precipitation regimes induced by elevation gradients (particularly in aspect‐controlled semi‐arid ecosystems) on landform patterns trying to understand feedbacks and consequences for coevolving vegetation has been limited. In this study, the Channel‐Hillslope Integrated Landscape Development (CHILD) landscape evolution model (LEM) coupled with the vegetation dynamics Bucket Grassland Model (BGM) is used to analyse the coevolution of semi‐arid landform‐vegetation ecosystems. The CHILD+BGM model is run under different combinations of precipitation and solar radiation settings. Three precipitation settings, including uniform, elevation control, and orographic control on precipitation, are considered in combination with spatially uniform and spatially varied radiation settings. Based on the results, elevation control, aspect, and drainage network are identified as the major drivers of the distribution of vegetation cover on the landscapes. Further, the combination of orographic precipitation and spatially varied solar radiation created the highest asymmetry in the landscape and divide migration due to the emergence of gentler slopes on the windward than the leeward sides of the domain. The modelling outcomes from this study indicate that aspect control of solar radiation in combination with orographic precipitation plays a key role in the generation of topographic asymmetry in semi‐arid ecosystems.
... Efficiency of the stabilizing effect of biocrust plus vegetation will depend on the intensity of abiotic factors controlling slope processes. Thus, at least solar radiation, contributing area and channel gradient, widely considered main drivers in slope processes (see, among others, Poulos et al., 2012;Richardson, 2015;Richardson et al., 2020), along with biocrusts and vascular vegetation (hereafter, vegetation) Roering et al., 2010;Yetemen et al., 2015b) should be considered as important drivers explaining asymmetry. ...
... At El Cautivo, solar radiation, contributing area, and channel slope were also important factors for NDAI. Except for biocrusts, all these factors are known to be main drivers affecting hillslope processes Poulos et al., 2012;Richardson, 2015;Richardson et al., 2020;Roering et al., 2010;Yetemen et al., 2015b). The findings presented here agree with Melton (1960) as asymmetries occurred at intermediate values of channel gradient, as well as with low order channels (as stated by Johnstone et al., 2017). ...
... Our field data confirmed differences in surface temperature and firmness. Thus, radiation-generated microclimatic divergence is a driver of the spatial organization of ecohydrologic fluxes, soil and vegetation patterns and dynamics, and landscape morphology in semiarid ecosystems Yetemen et al., 2015b). ...
Article
Full-text available
Catchment asymmetry is a fairly frequent phenomenon on a global scale but the main causes leading to its formation are still not well understood. Where the intervention of structural or tectonic causes is not relevant, asymmetry seems to result from differential erosion between opposite slopes that flow into the same channel, which is frequently associated with contrasted biocrust and/or vegetation covers. Biocrusts are known to be important surface stabilizing agents. However, their geomorphological consequences at the landscape scale are little known. In this study we combined field measurements with digital elevation models and image analysis to determine whether catchment asymmetry in the Tabernas Desert (semi-arid SE of Spain) is a local or general phenomenon, and to explore the main factors determining asymmetry occurrence, magnitude and direction. We pay special attention to the role of biocrusts. We found that catchment asymmetry is a very common phenomenon in the area; only 25% of the catchments are symmetrical, while approximately 40% present asymmetry with the relatively shady hillslope having a lower gradient, and 35% with that hillslope being the steeper. Solar radiation reaching the soil, surface area and channel gradient in the considered catchment stretch, as well as the total catchment area upstream from the lower point of the considered stretch were the main abiotic factors controlling the formation of the asymmetry. Microclimatic differentiation due to differences in radiation input caused by the uneven topography favoured the relative stabilization of the shadier hillslope and its colonization by biocrusts and later by plants. The effect of the biocrusts and vegetation protection against water erosion on shadier hillslopes is often stronger than that of the set of abiotic factors and gives rise to asymmetries with lower gradients in the shady hillslope by promoting lateral displacement of the channel. We hypothesised that the opposite pattern, with the sunnier hillslope having a lower gradient, occurs when abiotic factors control the development of asymmetry formation. In these conditions, the effect of biocrusts and plants would act in the opposite direction. We propose a conceptual model of feedbacks generating catchment asymmetry, with biocrust playing a crucial role.
... 44 Therefore, modifications on this variable surely provide metabolic variations, as remarked in several recent environmental studies 31,45 that are indicative of landscape modifications along natural history. 33,46 Radiation and temperature must correlate since sun radiation warms Earth, although many parameters affect the relationship between the above factors, 47 a fact that is consistent with the observations in Figure 5. In addition, this environmental factor can cause adaptative phenomena as that proposed by Charles Darwin,48 concluding that "effects of light (in some plants) do not correspond with its intensity" striving in physical adaptative variations that are visible and remarked in recent observations. ...
Article
The occurrence of racemic and enantiomerically enriched (scalemic) mixtures of secondary metabolites in their natural sources is a rare phenomenon. The unprecedent case of enantiomeric variations from levorotatory to dextrorotatory, and back to levorotatory, passing through an almost racemic mixture, was recently documented for areolal, the major epoxythymol of Piptothrix areolare. In an attempt to shed some light to understand the reasons for such an unusual behavior, herein, we evaluated this phenomenon by correlating the areolal enantiomeric purity with several environmental variables, including temperature, humidity, rain precipitation, wind speed, and radiation during over 1 year of the plant life cycle. The specific rotation and enantiomeric excess determined by 1H‐NMR‐BINOL measurements provided the scalemic variations of areolal samples isolated from the roots collected from the same location along a 427‐day period. The 1H‐NMR‐BINOL methodology provided better sensitivity to enantiomeric variations than specific rotation measurements. Statistical data, including matrix correlation analysis, exploratory analysis by heatmap plotting, and the principal component analysis (PCA), suggested direct correlation of the scalemic variation with humidity, rain precipitation, and radiation variables with the best PCA explanation (78.4%) and noncritical or poor correlations in PCA explained in 60.2% and 48.4%, respectively. When variations in the optical activity parameter of any metabolite are observed, the search for scalemic mixtures along their host plant life cycle should be undertaken. Herein, this phenomenon could be associated with interactions with soil microorganisms and with evolutionary aspects of Piptothrix areolare which belongs to Asteraceae, one of the most successfully adaptable plant families.
... Această categorie de factori derivă atât din conexiunile sistemice ale solurilor cu alte tipuri de ecosisteme din cadrul unui pedopeisaj, cât şi din condiţionările socio-economice, istorice sau culturale exercitate asupra resurselor de sol: (i) poluarea cu toate faţele ei (Charman, 1991;Verstraeten et al., 2002;Poch şi Martiénez-Casasnovas, 2006;Pollard şi Kibblewhite, 2006;Manandhar şi Odeh, 2014: Bruulsema, 2018; (ii) expansiunea urbană (în special, atunci când se manifestă în mod haotic şi necontrolat) (Hooke et al., 2012;Lehmann şi Stahr, 2007;Louwagie et. al., 2011;Panagos et al., 2016;Ferreira, 2018); (iii) gestionarea inadecvată şi defectuoasă a resurselor de apă (Eswaran et al., 2001;Montgomery, 2007;Yetemen et al., 2015); (iv) modificările climatice majore (Montanarella, 2017;IPCC, 2012IPCC, , 2014EEA, 2015;Thomas şi Lopez, 2015); (v) o serie de calamităţi şi dezastre naturale (inundaţii, alunecări de teren, incendii de vegetaţie necontrolate, erupţii vulcanice etc.) (Charman, 1991;Müller-Fürstenberger şi Schumacher, 2015) şi antropice (exploatarea fondului forestier, accidente industriale, conflicte militare etc.) (Hooke et al., 2012;Rusu et al., 2018;Francos et al., 2018); (vi) limitarea sau pierderea accesibilităţii oamenilor la resursele de sol (Aveyard şi Charman, 1991;FAO, 1995FAO, , 1997Poch şi Martiénez-Casasnovas, 2006;Krasilnikov et al., 2016); (vii) limitarea utilizării sustenabile a terenurilor indusă de factori socio-economici, culturali sau istorici etc. (De Alba et al., 2004;Kuhlman et al., 2010;Lanz et al., 2016;Brausmann şi Bretschger, 2018). ...
Chapter
Full-text available
At this moment we do not have a generally accepted terminology to reflect the classification of factors and processes of soil degradation. The present study approaches a series of aspects regarding this problem, with accents on the casual relations between factors, process dynamics an the impacts on soil, environment and human society.
... Therefore, the soil moisture response to rainfall events is more rapid in the model compared with the observed response, because the temporal characteristic of each rainfall event is not considered and water redistribution in soil is not considered in the model. The model captures the magnitudes of soil moisture pulses and the shapes of the soil-drying process in the Front and Middle reasonably well (Yetemen et al., 2015), although it underestimates the soil moisture in the summer as it overestimates evapotranspiration. However, the model performed less satisfactorily at the Back and the Control zones, possibly because the real meteorological conditions at these two zones are more complicated than are considered in the model, and there are obvious soil crusts in the Control which might raise the rainfall interception loss. ...
Article
Photovoltaic technology plays an important role in the sustainable development of clean energy, and arid areas are particularly ideal locations to build large-scale solar farms, all over the world. Modifications to the energy balance and water availability through the installation of large-scale solar farms, however, fundamentally affect the energy budget, water, and biogeochemical cycles. In-situ field observations, though, fail to draw definitive conclusions on how photovoltaic panels (PVs) affect the ambient environment, or how microclimates and soil moisture evolve under the long-term, continuous, cumulative influence of PVs. Here, we designed a synthetic model, integrating processes of energy budget and water cycle, to quantify the ecohydrological effects of PVs on soil microclimate and moisture regimes at different locations (zones) near individual PVs. Simulations run with a stochastically generated 100-year climate time series were examined to capture the evolutionary trends of soil microclimate and soil moisture. The results indicate that soil moisture content was increased by 59.8% to 113.6% in the Middle and Front zones, and soil temperature was decreased by 1.47 to 1.66 °C in all the sheltered zones, mainly because there was 5– 7 times more available water and ~27% less available radiation there, compared with the control zone. On the other hand, if the ground clearance of the PVs is too low, turbulence beneath hot PVs will have a significant influence on not only soil temperature but also soil moisture content. The innovative contribution of this study lies in reinforcing existing theoretical patterns for the development of soil microclimate and moisture dynamics influenced by PVs, and can be used to provide reliable insights into the hydrological and biogeochemical processes on Earth and the sustainable management of large-scale solar farms in arid ecosystems.
... The growth and dynamics of vegetation are affected by many environmental factors, such as microclimate, soil structure, water, solar radiation, and topography [6][7][8][9]. Among these factors, water is responsible for the growth and changes in desert plants, and directly affects the spatial and temporal distribution patterns of desert vegetation [10][11][12][13]. ...
Article
Full-text available
Surface water is an important factor affecting vegetation change in desert areas. However, little research has been conducted on the effects of surface water on vegetation expansion. In this study, the annual spatial distribution range of vegetation and surface water in the Daliyabuyi Oasis from 1990 to 2020 was extracted using Landsat time-series images. Based on multi-temporal and multi-scale remote sensing images, several plots were selected to demonstrate the process of landform change and vegetation expansion, and the influence of surface water on vegetation expansion was analyzed. The results show that the vegetation distribution and surface water coverage have increased from 1990 to 2020; and surface water is a critical factor that drives the expansion of vegetation. On the one hand, surface water in the study area was essential for reshaping the riparian landform, driving the transformation of dunes into floodplains, and increasing the potential colonization sites for vegetation. However, landform changes ultimately changed the redistribution of surface water, ensuring that enough water and nutrients provided by sediment were available for plant growth. Our study provides a critical reference for the restoration of desert vegetation and the sustainable development of oases.
... In reality, vegetation has played a significant role in shaping the Earth's terrestrial surface. Incorporating vegetation dynamics could make co-evolution results much more complicated, as their feedback mechanisms have a complex link with water, solar radiation, nutrient, carbon, sediment (soil), topography, etc. Modeling vegetation dynamics in the whole landscape context has gradually progressed in the last decade (e.g., McGrath et al., 2012;Yetemen et al., 2015). These efforts promise that vegetation dynamics can be eventually incorporated into the proposed co-evolution framework. ...
Article
Full-text available
Landscape evolution models simulate the long-term variation of topography under given rainfall scenarios. In reality, local rainfall is largely affected by topography, implying that surface topography and local climate evolve together. Herein, we develop a numerical simulation model for the evolution of the topography–climate coupled system. We investigate how simulated topography and rain field vary between “no-feedback” and “co-evolution” simulations. Co-evolution simulations produced results significantly different from those of no-feedback simulations, as illustrated by transects and time evolution in rainfall excess among others. We show that the evolving system keeps climatic and geomorphic footprints in asymmetric transects and local relief. We investigate the roles of the wind speed and the time lags between hydrometeor formation and rainfall (called the delay time) in the co-evolution. While their combined effects were thought to be represented by the non-dimensional delay time, we demonstrate that the evolution of the coupled system can be more complicated than previously thought. The channel concavity on the windward side becomes lower as the imposed wind speed or the delay time grows. This tendency is explained with the effect of generated spatial rainfall distribution on the area–runoff relationship.
Article
Biomass of seedlings at different developing stages of growth is important information for studying the response of species to site conditions. The objectives of this study was to explore the distribution characteristics of AGB (above-ground biomass) and BGB (below-ground biomass) of Abies georgei var. smithii seedlings of different ages, and investigate the effects of topography (slope aspect, altitude), plant community characteristics (crown density, species diversity, etc.), and soil properties (soil physical and chemical properties) on the biomass and its allocation. Seedlings in five age classes (1–2, 3–4, 5–6, 7–8, and 9–10 years old) were collected by full excavation from 6 elevations (3800 m, 3900 m, 4000 m, 4100 m, 4200 m, 4300 m) on the north and south slopes of Sejila Mountain in Tibet. 15 seedlings of each age class were investigated at one altitude. The individual effects of seedling age (SA) and the interaction effects of SA, slope aspect (SL), and elevation (EG), namely, SL×EG, SL×SA, EG×SA, and SL×EG×SA, had significant effects on the AGB of the seedlings (p<0.05), whereas BGB was only significantly affected by SA (p<0.001). The AGB and BGB of the seedlings showed a binomial growth trend with the increase in seedling age, and had an allometric relationship at different elevations, α (allometric exponential) varied from 0.913 to 1.046 in the northern slope, and from 1.004 to 1.268 in the southern slope. The biomass of seedlings on the northern slope was remarkably affected by stand factors, with a contribution rate of 47.8%, whereas that on the southern slope was considerably affected by soil factors with a contribution rate of 53.2%. The results showed that age was the most important factor affecting seedling biomass. The allometric pattern of seedling biomass was relatively stable, but in a high-altitude habitat, A. georgei var. smithii seedlings increased the input of BGB. Understanding seedling biomass allocation and its influencing factors is useful for evaluating plants’ ability to acquire resources and survival strategies for adaptation to the environment in Tibet Plateau.
Article
Water stored in soils, in part, controls vegetation productivity and the duration of growing seasons in wildland ecosystems. Soil water is the dynamic product of precipitation, evapotranspiration, and soil properties, all of which vary across complex terrain making it challenging to decipher the specific controls that soil water has on growing season dynamics. We assess how soil water use by plants varies across elevations and aspects in the Dry Creek Experimental Watershed in southwest Idaho, USA, a mountainous, semiarid catchment that spans low elevation rain to high elevation snow regimes. We compare trends in soil water and soil temperature with corresponding trends in insolation, precipitation, and vegetation productivity, and we observe trends in the timing, rate, and duration of soil water extraction by plants across ranges in elevation and aspect. The initiation of growth-supporting conditions, indicated by soil warming, occurs 58 days earlier at lower, compared to higher, elevations. However, growth-supporting conditions also end earlier at lower elevations due to the onset of soil water depletion 29 days earlier than at higher elevations. A corresponding shift in peak NDVI timing occurs 61 days earlier at lower elevations. Differences in timing also occur with aspect, with most threshold timings varying by 14-30 days for paired north- and south-facing sites at similar elevations. While net primary productivity nearly doubles at higher elevations, the duration of the warm-wet period of active water use does not vary systematically with elevation. Instead, the greater ecosystem productivity is related to increased soil water storage capacity, which supports faster soil water use and growth rates near the summer solstice and peak insolation. Larger soil water storage does not appear to extend the duration of the growing season, but rather supports higher growing season intensity when wet-warm soil conditions align with high insolation. These observations highlight the influence of soil water storage capacity in dictating ecological function in these semiarid steppe climatic regimes. This article is protected by copyright. All rights reserved.
Chapter
Full-text available
This chapter reviews quantitative modeling of landscape evolution. Quantitative modeling is contrasted with conceptual or physical modeling, and four categories of model studies are presented. Procedural studies focus on model experimentation. Descriptive studies use models to learn about landscapes in general. Postdictive and predictive try to correctly simulate the evolution of real landscapes, respectively in the past (with calibration) or in the future (with calibrated models). After classification of 322 landscape evolution studies in these categories, we find that descriptive studies are most common, and predictive studies are least common. Procedural studies have focused on production methods for digital landscapes, spatial resolution effects, the role of sinks and depressions and calculation schemes for flow routing. Descriptive studies focused mainly on surface-tectonic interactions, sensitivity to external forcing, and the definition of crucial field observations from model results. Postdictive and predictive studies operate mainly in time-forward mode and are increasingly validated using independent data. Overall, landscape evolution modeling has progressed to the extent that non-experts are able to easily use modern models, and are commonly used in inversion schemes to obtain the most likely (set of) inputs to produce known topographies. This development will likely continue, with more attention for interactions with ecology and soils over short (ca, ma) timescales, and with climate over long (Ma) timescales.