The Evaporative Demand Drought Index: Part II – CONUS-wide Assessment Against Common Drought Indicators

Abstract

Precipitation, soil moisture, and air temperature are the most commonly used climate variables to monitor
drought; however, other climatic factors such as solar radiation, wind speed, and humidity can be important
drivers in the depletion of soil moisture and evolution and persistence of drought. This work assesses the
Evaporative Demand Drought Index (EDDI) at multiple time scales for several hydroclimates as the second
part of a two-part study. EDDI and individual evaporative demand components were examined as they relate
to the dynamic evolution of flash drought over the central United States, characterization of hydrologic
drought over the western United States, and comparison to commonly used drought metrics of the U.S.
Drought Monitor (USDM), Standardized Precipitation Index (SPI), Standardized Soil Moisture Index (SSI),
and the evaporative stress index (ESI). Two main advantages of EDDI over other drought indices are that it
is independent of precipitation (similar to ESI) and it can be decomposed to identify the role individual
evaporative drivers have on drought onset and persistence. At short time scales, spatial distributions and time
series results illustrate that EDDI often indicates drought onset well in advance of the USDM, SPI, and SSI.
Results illustrate the benefits of physically based evaporative demand estimates and demonstrate EDDI’s
utility and effectiveness in an easy-to-implement agricultural early warning and long-term hydrologic
drought–monitoring tool with potential applications in seasonal forecasting and fire-weather monitoring

Full text

The Evaporative Demand Drought Index. Part II: CONUS-Wide Assessment
against Common Drought Indicators
DANIEL J. MCEVOY,
a
JUSTIN L. HUNTINGTON,
a
MICHAEL T. HOBBINS,
b,c
ANDREW WOOD,
d
CHARLES MORTON,
a
MARTHA ANDERSON,
e
AND CHRISTOPHER HAIN
f
a
Desert Research Institute, Reno, Nevada
b
Cooperative Institute for Research in Environmental Sciences, University of Colorado Boulder, Boulder, Colorado
c
NOAA/Earth Systems Research Laboratory/Physical Sciences Division, Boulder, Colorado
d
National Center for Atmospheric Research, Boulder, Colorado
e
Hydrology and Remote Sensing Laboratory, Agricultural Research Service, U.S. Department of Agriculture, Beltsville, Maryland
f
Earth System Science Interdisciplinary Center, University of Maryland, College Park, College Park, Maryland
(Manuscript received 22 July 2015, in final form 18 April 2016)
ABSTRACT
Precipitation, soil moisture, and air temperature are the most commonly used climate variables to monitor
drought; however, other climatic factors such as solar radiation, wind speed, and humidity can be important
drivers in the depletion of soil moisture and evolution and persistence of drought. This work assesses the
Evaporative Demand Drought Index (EDDI) at multiple time scales for several hydroclimates as the second
part of a two-part study. EDDI and individual evaporative demand components were examined as they relate
to the dynamic evolution of flash drought over the central United States, characterization of hydrologic
drought over the western United States, and comparison to commonly used drought metrics of the U.S.
Drought Monitor (USDM), Standardized Precipitation Index (SPI), Standardized Soil Moisture Index (SSI),
and the evaporative stress index (ESI). Two main advantages of EDDI over other drought indices are that it
is independent of precipitation (similar to ESI) and it can be decomposed to identify the role individual
evaporative drivers have on drought onset and persistence. At short time scales, spatial distributions and time
series results illustrate that EDDI often indicates drought onset well in advance of the USDM, SPI, and SSI.
Results illustrate the benefits of physically based evaporative demand estimates and demonstrate EDDI’s
utility and effectiveness in an easy-to-implement agricultural early warning and long-term hydrologic
drought–monitoring tool with potential applications in seasonal forecasting and fire-weather monitoring.
1. Introduction
Drought is a complex and naturally occurring process
with adverse effects on society, primarily through deg-
radation and loss of agricultural crops and depletion of
water resources (i.e., streamflow and reservoir storage).
Recent examples are instructive: in California, the ex-
tended drought that began in late 2011 is still ongoing,
and the 2011–14 three-year average precipitation (Prcp)
record indicates that this period is the second driest in
recorded history (Seager et al. 2015). In 2011, Texas
experienced extreme Prcp deficits, while in 2011 and
2012 record-breaking air temperatures T
air
and high
wind speeds Uplayed a significant role in drought in-
tensification over much of the central United States
(Karl et al. 2012;Cattiaux and Yiou 2013). Total eco-
nomic losses are estimated to be $2.7 billion, $7.7 billion,
and more than $35 billion for the California, Texas, and
central U.S. droughts, respectively. While conditions in
Texas deteriorated over many months in 2011, the de-
pletion of moisture over the central United States in
2011 occurred at a much faster rate. This fast onset of
drought has recently been termed ‘‘flash drought’’
(Svoboda et al. 2002). The physical mechanisms driving
flash droughts have been largely neglected from tradi-
tional drought metrics. Hence, there is a growing need
for continued development of physically based drought
metrics that capture Prcp-independent land surface–
atmosphere feedbacks, specifically the complementary
relationship between actual evapotranspiration (ET)
and atmospheric evaporative demand E
0
.
Corresponding author address: Daniel J. McEvoy, Western Re-
gional Climate Center, Desert Research Institute, 2215 Raggio
Parkway, Reno, NV 89512.
E-mail: mcevoyd@dri.edu
JUNE 2016 M C E V O Y E T A L . 1763
DOI: 10.1175/JHM-D-15-0122.1
Ó2016 American Meteorological Society

It has been common practice in recent decades to
monitor and analyze drought using metrics driven by
Prcp and T
air
only. The two most commonly used
drought indices are the Palmer drought severity index
(PDSI; Palmer 1965), which relies on monthly T
air
and Prcp, and the Standardized Precipitation Index
(SPI; McKee et al. 1993), which relies on Prcp only.
While the PDSI and SPI have proven useful for pro-
viding valuable information regarding hydrologic and
meteorological drought, these metrics have limitations
at short time scales and fail to account for the effects of
other important meteorological and radiative forcings
such as specific humidity q,U, and downwelling short-
wave radiation R
d
. The most heavily used dataset for
decision-making with regards to drought is the U.S.
Drought Monitor (USDM; Svoboda et al. 2002), which
relies on a blend of metrics (including PDSI and SPI) and
hydrologic data [e.g., soil moisture (SM), streamflow, and
snow water equivalent] to produce weekly maps of
drought severity. The USDM could be improved through
the inclusion of important hydrometeorological forcings
key to identifying flash and long-term drought through
the use of physically based E
0
estimates.
Other operational products could similarly be im-
proved with the inclusion of physically based E
0
estimates. For example, the U.S. operational PDSI,
produced by the National Oceanic and Atmospheric
Administration (Heddinghaus and Sabol 1991), con-
tinues to use T
air
-based E
0
estimates (i.e., Thornthwaite
1948) within the PDSI formulation despite the fact that
there have been a number of studies that recommend
the use of physically based formulations of E
0
(e.g.,
Palmer 1965;Jensen 1973;Hobbins et al. 2008,2012;
Roderick et al. 2009;Milly and Dunne 2011;Hobbins
2016). Both Dai (2011) and van der Schrier et al. (2011)
found PDSI to be largely insensitive to E
0
parameteri-
zation during the twentieth and early twenty-first centu-
ries. On the other hand, Sheffield et al. (2012) found
major differences between the PDSI driven with T
air
-based
and physically based E
0
estimates, especially from the mid-
1990s through 2008, with T
air
-based E
0
estimates sh owi ng a
significant drying trend and physically based E
0
es-
timates indicating no significant trend in global
drought severity. The role of physically based E
0
es-
timates in drought monitoring and prediction remains
an active area of research and is a focus of this paper.
Recent studies have shown that ET, which is obtained
through the use of thermal and optical satellite remote
sensing or land surface models, used in combination
with physically based E
0
, can be used as a drought in-
dicator by inherently accounting for feedbacks at the
land surface–atmosphere interface through the use of
ratios of ET to E
0
(Anderson et al. 2007a,b,2011;Yao
et al. 2010;Mu et al. 2013;Otkin et al. 2013,2014).
Evapotranspiration-based drought indices that use op-
tical and thermal remote sensing, such as the evapora-
tive stress index (ESI; Anderson et al. 2007a,b,2011),
have the advantage of being sensitive to rapid changes
in soil moisture conditions that are driven by changes in
the atmospheric drivers of T
air
,U,q, and R
d
and the
unique ability to provide early warning of flash drought
development (Otkin et al. 2013). Some of the limita-
tions of using remotely sensed drought indices include
cloud cover, satellite interarrival times that have to be
interpolated, and limited record length for a robust
climatology.
To complement indices like ESI, which estimate ac-
tual stress on the ground experienced by the vegetation,
Hobbins et al. (2016, hereafter Part I) developed the
Evaporative Demand Drought Index (EDDI), a mea-
sure of the drying potential of the atmosphere that can
presage vegetative stress on the ground. Part I describes
two primary physical feedbacks between ET and E
0
that form the rationale for EDDI: a complementary
relationship under water-limited conditions (extended
drought) where ET and E
0
vary in opposite directions
(Bouchet 1963) and a parallel relationship under
energy-limited conditions at the onset of flash drought.
Under both scenarios, EDDI was found to respond to
drying and wetting anomalies of major components of
the hydrologic cycle (streamflow, ET, Prcp, and SM) at
monthly to annual time scales in several river basins
over the contiguous United States (CONUS) with dif-
ferent hydroclimates (Part I). At flash drought time
scales (from weekly to monthly), increased E
0
(high
EDDI) may not always lead to vegetative stress. To
confirm drought stress, a thermal infrared–based remote
sensing approach such as ESI can be useful.
This paper builds upon the work of Part I through a
CONUS-wide assessment of EDDI against several
commonly used drought indices. Data sources and
methodology are presented first, followed by compari-
sons of EDDI to other commonly used drought metrics,
flash drought case studies over the central and north-
eastern United States, and finally, extended drought
case studies over the western United States.
2. Data and methods
a. Evaporative demand
Various methods have been developed to compute E
0
,
including T
air
-based methods (e.g., Thornthwaite 1948;
Hargreaves and Samani 1985), radiation-based methods
(Priestley and Taylor 1972), and radiation–aerodynamic
combination methods that incorporate maximum tem-
perature T
max
, minimum temperature T
min
,R
d
,U, and
1764 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

q, such as the Penman–Monteith (PM) approach (Monteith
1965). A priori, it is generally assumed that if the nec-
essary data resources are available, a full-form physi-
cally based method, such as PM, should be used over
methods based only on T
air
and/or radiation. Hobbins
et al. (2012) and Hobbins (2016) demonstrated that the
primary drivers of E
0
variability differ across the United
States and with aggregation period (e.g., monthly vs annual)
and season. For example, during summer months Uis
the primary driver of E
0
variability over much of the
Great Basin, while R
d
is the primary driver of vari-
ability over much of the southeastern United States. In
this study, we use reference ET from the PM-based
American Society of Civil Engineers standardized ref-
erence ET equation (Allen et al. 2005)forE
0
.Daily
bias-corrected and spatially disaggregated (from 12 to
4 km) gridded surface meteorological data (METDATA;
Abatzoglou 2013) are used to compute E
0
on a daily basis
for 1979–2015. Variables T
max
,T
min
,qat 2 m, R
d
,andU
[adjusted from 10 to 2 m following Allen et al. (2005)]
were obtained from the University of Idaho (http://
metdata.northwestknowledge.net/).
b. Evaporative Demand Drought Index
A probability-based standardized climate variable
can be obtained using parametric or nonparametric
methods. Parametric methods use a single probability
distribution to fit a time series (e.g., Gamma distribution
for SPI), where probabilities are transformed to stan-
dardized values through an inverse normal approxima-
tion. However, a single probability distribution may
not always be appropriate at large spatial scales, and
several studies have documented these limitations with
SPI (Guttman 1999;Quiring 2009) and Standardized
Streamflow Index (Vicente-Serrano et al. 2012). The
EDDI calculation procedure is presented in Part I and
uses a nonparametric probability-based approach to
allow for more consistent comparisons between EDDI
and other standardized indices.
The EDDI methodology follows Hao and AghaKouchak
(2014), where the plotting position approach was used
to compute SPI, Standardized Soil Moisture Index
(SSI), and Multivariate Standardized Drought Index
(MSDI). Farahmand and AghaKouchak (2015) rec-
ommend this plotting position approach to maintain con-
sistency when comparing several standardized drought
indices.
c. Comparison drought metrics
1) NLDAS-2-BASED DROUGHT INDICES
To assess the ability of EDDI to identify historical
drought periods, EDDI is compared to SPI and SSI
using Prcp and simulated SM from the North American
Land Data Assimilation System phase 2 (NLDAS-2;
Xia et al. 2012a,b). NLDAS-2 Prcp is primarily de-
rived from Climate Prediction Center gridded daily
gauge data [with a topographic adjustment from the
Parameter-Elevation Regressions on Independent
Slopes Model (PRISM; Daly et al. 1994)]. NLDAS-2
SM is derived from the Variable Infiltration Capacity
model (VIC; Liang et al. 1994) and represents the av-
erage SM from the top 100 cm of the soil column. Daily
NLDAS-2 data (Xia et al. 2012a,b) were provided
(courtesy of Youlong Xia, NCEP) and used only for
time series analysis of flash drought case studies (Fig. 8,
described in greater detail below). Monthly NLDAS-2
data were obtained for the period of 1979–2013 with a
native grid spacing of 0.1258. To compare EDDI to
NLDAS-2 drought indices, all NLDAS-2 data were
resampled to the 4-km (;1
/
168) METDATA grid using
bilinear interpolation. Precipitation and SM were ac-
cumulated at five time scales (1, 3, 6, 9, and 12 months)
and standardized following the EDDI methodology of
plotting positions and inverse normal approximation
(Part I). Pearson linear correlation coefficients between
EDDI and standardized NLDAS-2 drought indices were
computed for each month (n535 years) at the five
time scales.
2) EVAPORATIVE STRESS INDEX
The ESI (Anderson et al. 2007b,2011) represents
standardized anomalies in the ET fraction of reference
ET (i.e., ET/E
0
), with ET obtained through satellite-
assisted modeling of the land surface energy balance.
Evapotranspiration and other land surface energy bal-
ance components are retrieved using satellite optical
and thermal imagery to force the Atmosphere–Land
Exchange Inverse (ALEXI) surface energy balance
model (Anderson et al. 1997,2007a). Atmospheric var-
iables needed to drive ALEXI come from the North
American Regional Reanalysis (NARR; Mesinger
et al. 2006).
Weekly ESI data were provided over the United
States for 2000–13 at a 4-km spatial resolution and
were aggregated to time scales of 1, 2, and 3 months.
To obtain a consistent comparison between EDDI
and ESI, EDDI was recalculated using the same pe-
riod of record as the ESI (n514 years) and the same
aggregation time scales. ESI data were resampled
using bilinear interpolation to match the EDDI grid.
No downscaling was necessary as both grids were of
identical spatial resolution. Pearson linear correlation
coefficients between EDDI and ESI were computed
for each week over the 14-yr period and at all five
time scales.
JUNE 2016 M C E V O Y E T A L . 1765

3) U.S. DROUGHT MONITOR
The USDM (Svoboda et al. 2002) was used as another
metric to assess EDDI, with the primary goal of identi-
fying differences between the two metrics during the
evolution of drought through time and space. The
USDM is derived from a blend of drought metrics ad-
justed using local expert knowledge to develop weekly
drought severity maps over CONUS (Svoboda et al.
2002;Anderson et al. 2013). The USDM classification
system of drought ranges from D0 (abnormally dry) to
D4 (exceptional drought). For results where the USDM
is compared, all drought metrics were converted to
USDM classes (Table 1). The comparisons of EDDI to
the USDM are necessarily qualitative because the USDM
is a blend of information at several different time scales,
whereas EDDI represents a single time scale.
USDM data (2000–13) were downloaded as Envi-
ronmental Systems Research Institute, Inc. (ESRI),
shapefiles provided by the National Drought Mitigation
Center and rasterized to match the 4-km EDDI grid to
create a USDM class map of integer values of drought
intensity ranging from 0 to 4 (i.e., D050, D151, D252,
D353, and D454).
3. Results
a. NLDAS-2 drought index correlations with EDDI
Temporal correlations rbetween EDDI and NLDAS-2
drought indices (EDDI–SPI and EDDI–SSI) for 1-, 6-,
and 12-month time scales are shown in Fig. 1. Drought
potential and drought itself are indicated by positive
EDDI values and negative SPI and SSI values; there-
fore, strong negative correlations represent similar
drought signals between EDDI and both SPI and SSI
over the 35-yr period of record. At the 1–12-month time
scales, correlations between EDDI and SPI and SSI are
strongest (more negative) over much of the southwest-
ern and south-central United States (with the exception
of 1-month SSI), and highest in Texas (r,20.7). The
Northeast is a region of general weak correlations for
both EDDI–SPI and EDDI–SSI, with the Midwestern
states of Ohio, Indiana, and Michigan being a weak
spot for EDDI–SPI only. Spatial correlations at 6- and
12-month time scales are quite similar (Figs. 1c–f) and
are generally much stronger than at the 1-month time
scale (Figs. 1a,b). Over the northeastern United States,
EDDI–SPI correlations remain fairly weak at longer
time scales, while EDDI–SSI correlations improve over
Ohio, West Virginia, New York, and Pennsylvania
(Figs. 1c–f).
Weak correlations to 1-month SSI over the western
United States may be explained by above-average T
air
and R
d
(driving EDDI upward) that can lead to in-
creased snowmelt and SM, and a short-term wetting
signal from SSI, particularly during the winter months.
Positive correlations of EDDI–SPI and EDDI–SSI over
the northeastern United States are caused by energy-
limited conditions, as opposed to water-limited condi-
tions. In such regions, the rate of change in ET is
generally proportional and in the same direction as
E
0
(Han et al. 2014;Part I).
Figure 2 highlights four regions of interest selected for
individual monthly correlation analysis. The Central
Valley of California and Iowa are two major agricultural
regions where drought impacts can have adverse effects
on crop production. East-central Texas is part of a re-
gion that has been identified as a global ‘‘hot spot’’ for
strong land surface–atmosphere coupling (Koster et al.
2004,2006); therefore, strong correlation of SM and
Prcp to EDDI is expected. Koster et al. (2009) identified
Pennsylvania as an area where generally high SM exerts
little control on ET because of prevailing energy-
limiting conditions, even during times of severe meteo-
rological drought. This observation is consistent with
low correlations found in Fig. 1 in parts of the north-
eastern United States. The following section further
highlights how E
0
anomalies (i.e., EDDI) in Pennsyl-
vania relate to SM- and Prcp-driven droughts.
Individual monthly correlations between EDDI and
NLDAS-2-derived indices at various time scales are
shown in Fig. 3 for these regions of interest. For each of
the selected regions shown in Fig. 2, EDDI correlations
to SSI and SPI were area-averaged over all pixels. For
TABLE 1. Drought classes for comparing USDM to SPI, SSI, ESI, and EDDI. Positive EDDI values indicate drought with the upper
percentiles (70–100) used to derive USDM classes.
USDM drought
category Description SPI, SSI, and ESI percentiles EDDI percentiles
D0 Abnormally dry 21–30 70–79
D1 Moderate drought 11–20 80–89
D2 Severe drought 6–10 90–94
D3 Extreme drought 3–5 95–97
D4 Exceptional drought 0–2 98–100
1766 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

the Texas region (Figs. 3a,e), seasonality and time scale
had little impact on the strength of correlations and
generally showed strong inverse relationships (r,20.6
for SPI and r,20.7 for SSI) during most months and
time scales, supporting the conclusions of Koster et al.
(2004,2006).
For the California region, large seasonal and time
scale–dependent variations were found, especially at the
1-month time scale for both SPI and SSI (Figs. 3b,f).
Correlations ranged from 10.20 to 20.82, with the
highest correlations occurring at the 6–12-month time
scales during the growing season. An exceptionally
weak correlation (20.13) was found with SPI during
July at the 1-month time scale. July is the driest month of
the year for the Central Valley of California, and most
Julys see zero Prcp accumulation. This limits the nega-
tive range of the 1-month SPI (McEvoy et al. 2012),
causing poor correlations with EDDI. Furthermore,
when it does rain during dry summer months, it occurs
from isolated convective activity over a single day: even
if most of the month is warm, cloud free, and dry
(leading to a drought signal from EDDI), the SPI
shows a wet anomaly. A more consistent stepped cor-
relation pattern was revealed at longer time scales,
where rvalues ,20.7 were found during the spring
(April–June) for 3-month periods, spring and summer
(July–September) for 6-month periods, and summer and
fall (October–December) for 9- and 12-month periods.
Iowa was similar to Texas in that little variability was
found in correlations (rvalues only ranged from 20.5
to 20.7), with the exception of the 1-month time scale.
Lower correlations at 1-month time scales during the fall
and winter should be expected with SSI, since the top
100 cm of ground is typically frozen during these months
and land surface–atmosphere coupling is weak. There
is a rapid increase in correlation at the 1-month time
scale during the late spring and summer.
Correlations for the Pennsylvania region were the
weakest of the four analyzed, with notably higher cor-
relations to SSI (Fig. 3h) than to SPI (Fig. 3d). EDDI is
FIG. 1. Correlation coefficients between EDDI and SPI at (a) 1-month, (c) 6-month, and (e) 12-month time scales
and between EDDI and SSI at (b) 1-month, (d) 6-month, and (f) 12-month time scales. Correlations were computed
at each grid point for 1979–2013 over each month (n535) and then averaged over all months in each time scale.
JUNE 2016 M C E V O Y E T A L . 1767

independent of Prcp and can be highly positive even
during times of Prcp surplus. However, EDDI is not
completely independent of SSI since changes in soil
moisture are partially controlled by the drivers of
physically based E
0
. For SPI (Fig. 3d), rvalues never
exceed 20.56, while for SSI (Fig. 3h)rvalues ranged
from 20.60 to 20.69 during the summer and early fall at
1-, 3-, and 6-month time scales. Weak correlations were
found to be both slightly positive and negative (20.30 ,
r,10.20) for SPI and SSI at the 1-month time scale
during fall and winter, and for winter and spring months
at other time scales. Results shown in Fig. 3 illustrate
that EDDI may be particularly useful for flash drought
and seasonal drought monitoring, especially during the
growing season.
b. ESI correlations with EDDI
Seasonal temporal correlations between ESI and
EDDI for CONUS are shown in Fig. 4. Only spring
(April–June) and summer (July–September) periods are
evaluated because of limited availability of continuous
monthly ESI data during fall and winter. ESI data were
frequently missing in snow-covered mountainous re-
gions of the west during spring and summer periods, and
ESI pixels were masked (indicated by white shading in
Fig. 4, as in the mountain ranges of western United
States) when less than 75% of the monthly time series
was available over the period of 2000–13. One benefit of
EDDI over ESI and other remote sensing–based metrics
is that EDDI can be used during all seasons. This may be
particularly useful for high-elevation hydrometeoro-
logical monitoring in seasonally snow-covered areas.
Figure 4 illustrates fairly large differences between
spring and summer periods, with negligible differences
between different time scales of 4, 8, and 12 weeks.
During the spring period, negative correlations are
strongest (r,20.7) over much of Texas, the desert
Southwest, and the Central Valley of California, while
weaker relationships were found over the Northeast and
parts of the Pacific Northwest (Figs. 4a,c,e). The low
positive correlations in the Northeast are due to energy-
limited evaporative conditions described in section 3a.
Summer correlations (Figs. 4b,d,f) are strongest, and
spatial patterns most consistent, over the central United
States, and lower correlations are evident over parts of
Nevada, California, and into the Pacific Northwest when
compared to the spring period. Low summer correla-
tions in Florida may be due to the shallow water table
enhancing actual ET, and EDDI may not be a good in-
dicator of drought potential in this region. Inspection of
the summertime series from the regions of low correla-
tion in the west and Pacific Northwest showed that,
during certain summers, ESI and EDDI were strongly
negatively correlated but positively correlated in others
(not shown). Evapotranspiration rates in semiarid re-
gions are typically low during summer periods; there-
fore, small variations in ET can potentially lead to large
changes in ESI, making for poor correlations with EDDI.
For example, most of Nevada experienced below-normal
Prcp and high temperatures for July 2005, and EDDI and
SPI indicated drought conditions, whereas ESI indicated
wet conditions (not shown). In general, EDDI is strongly
correlated to ESI (r,20.7) during spring and summer
months over much of the Southwest, south-central, and
north-central United States.
c. Flash drought during the growing season
Flash drought can develop even during periods of
normal or excess Prcp, and evaporative drivers can
uniquely identify the onset and evolution of flash drought.
For example, in some situations, a temperature-based E
0
would fail to identify rapid drying due to below-normal
T
air
coincident with high Uand low q. The following
highlights the Midwest droughts of 2011 and 2012 and
several other case studies in the central and northeastern
United States to demonstrate how EDDI can serve as an
effective early warning of flash droughts.
Area-averaged time series of 1-month EDDI are
compared to 1-month SPI and SSI during 2011 and 2012
in Fig. 5 for the Iowa domain. Note that the vertical axis
of EDDI is inverted to better visualize drought onset
and duration when compared to SPI and SSI in Fig. 5.
Figure 5 illustrates that in April 2011, all indices are
near neutral (i.e., close to zero), and over the next
2 months EDDI changes to a moderate drought class
(.0.78 or USDM D1 class), while both SPI and SSI in-
crease to slightly wet conditions. SPI and SSI values
FIG. 2. Case study areas. Shading indicates METDATA terrain
height (m) and red boxes indicate area-averaging domains for
Fig. 3. Iowa (IA), Texas (TX), and Pennsylvania (PA) boxes are
50 3100 4-km METDATA pixels (200 km 3400 km), and the
California (CA) box is 25 325 pixels (100 km 3100 km). Blue
patches indicate area-averaging domains for the flash drought cases
in Fig. 8.
1768 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

show no decrease until July 2011 (see black box in
Fig. 5). EDDI reaches D1 in June, while SSI and SPI
reach D1 in September, indicating a 3-month lead pro-
vided by EDDI. SPI falls below moderate drought in
September, and SSI follows 1 month later in October.
Both EDDI and SSI maintain extended drought condi-
tions throughout all of 2012, with the exception of
February, when EDDI is slightly above moderate
drought (0.78), but still below zero. During this ex-
tended drought of 2012, SPI is highly variable and in-
dicates wet conditions for many months.
To highlight the E
0
drivers that caused EDDI to signal
first a flash drought and then an extended drought, a
simple sensitivity analysis of EDDI was performed
(Figs. 6a,b). For this analysis, E
0
was calculated while
constraining the variable of interest to daily climatology
values in order to isolate the impact of each forcing
on the EDDI drought signal. Results are presented as
estimates of EDDI with a notation of the variable of
interest (i.e., EDDI-T,EDDI-q,EDDI-R
d
,andEDDI-U).
For example, EDDI-Twas calculated using the daily
climatology of T
max
and T
min
, and with METDATA-
observed values for all other variables. During the pe-
riod of 20–25 May 2011, EDDI-qand EDDI-Uhad the
greatest separation from standard EDDI values in the
negative direction (note yaxis is inverted), which in-
dicates that the drying power of the air term in the E
0
equation (Umultiplied by vapor pressure deficit) initi-
ated the flash drought signal in EDDI via increased U
and below-normal q(Fig. 6b) during the period from
20 May through 5 June. In this case, using daily clima-
tology qand Uvalues mitigated the drought signal rel-
ative to the standard EDDI. By June 2011, EDDI
decreased below the moderate drought threshold (0.78),
with the primary difference from May being that Uand
T
air
were then acting in combination to exacerbate the
FIG. 3. Monthly correlations between (top) EDDI and SPI and (bottom) EDDI and SSI at all time scales for (a),(e) central Texas; (b),(f)
the California Central Valley; (c),(g) Iowa; and (d),(h) Pennsylvania. The yaxis indicates ending month of each time scale, and the x
axis shows time scale (months). Shading indicates correlation coefficients. Correlations were computed at each grid point for 1979 to 2013
(n535) and then averaged over each climate region.
JUNE 2016 M C E V O Y E T A L . 1769

drought signal—as opposed to T
air
moderating it in May.
Despite below-normal T
air
conditions in September
2011 (Fig. 6a), the standard EDDI drought signal was
maintained because of extremely low qvalues evidenced
by a large difference between EDDI and EDDI-q(ab-
solute difference of 1.17). From November 2011 through
May 2012, T
air
dominated the EDDI signal, as seen by
the large differences between EDDI and EDDI-T.This
increase in T
air
and E
0
likely contributed to the per-
sistent SSI drought signal throughout 2012, despite
above-normal Prcp for February, April, October, and
December (see Fig. 5).
FIG. 5. EDDI under sustained and flash drought conditions. Monthly time series of 1-month
EDDI, SSI, and SPI area averaged over the Iowa domain for 2011 and 2012. Note that the yaxis
of EDDI is inverted to clearly visualize drought onset and duration relative to SPI and SSI.
Light green reference line indicates start of moderate drought classification (EDDI 50.78, SPI
and SSI 520.78).
FIG. 4. Seasonal (April–September) correlation coefficients [(left) spring and (right) summer] between ESI and
EDDI at (a),(b) 4-week; (c),(d) 8-week; and (e),(f) 12-week time scales. Areas in white indicate an insufficient
amount of ESI data. Correlations were computed at each grid point for 2000–13 over each week (n514 years) and
then averaged over all months.
1770 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

To spatially assess EDDI during the 2012 drought, a
comparison was made between the USDM, SPI, SSI,
and ESI. Recall from section 2e that the USDM is a
blend of products at various time scales, but we are
comparing it here to a fixed time scale EDDI: thus, the
EDDI and the USDM distributions should not be ex-
pected to be identical. The objective of the EDDI and
USDM comparisons is to show that EDDI can presage
rapid onset droughts before the impacts show up in the
USDM, thus highlighting the substantial added value
gained by using EDDI in conjunction with other drought-
monitoring metrics for decision-making applications.
Figure 7 shows the evolution of the 1-month EDDI,
ESI, SSI, and SPI, and USDM through time and space
over the spring and summer of 2012. The USDM gen-
erally indicated no drought or only D1–D2 over much
of the central United States as of 1 May (Fig. 7a). In
contrast, EDDI indicates at least moderate drought
conditions over most of the same region and looks
similar to the USDM spatial distribution of 2 months
later (i.e., of 3 July 2012). EDDI responded to anoma-
lously high T
air
,U, and R
d
across the region during the
second half of April. ESI showed widespread neutral
conditions for April with a rapid intensification in May.
SSI and SPI show a slower progression and more local
intensification (nonuniform spatial distribution) when
compared to EDDI and ESI. The 2012 drought evolu-
tion illustrated by the USDM over the central United
States expands in both spatial extent and severity
throughout the summer; however, the progression from
D0toD3 and D4 takes approximately 3 months.
Figure 7 illustrates that 1-month EDDI presaged the
onset of USDM extreme to exceptional drought (D3–
D4) by as much as 2 months. This case study highlights
the application of using EDDI to identify future drought
potential and onset of drought.
Four additional flash drought cases from Pennsylva-
nia, New York, Wisconsin, and the Ohio–Indiana border
are presented in Fig. 8 using daily time series of 1-month
EDDI, SPI, and SSI during the growing season (April–
September). Two of these cases (Wisconsin 2002 and
Ohio–Indiana 2007) are examples from Otkin et al.
(2014). Two new cases (Pennsylvania 1983 and New
York 1991) are presented in this study to show that
EDDI is effective in energy-limited regions such as
Pennsylvania, despite low correlations to SPI there
(see Fig. 3d). All four case studies are located in major
agricultural regions. Domains used for spatial averaging
FIG. 6. (a) Monthly time series (values on the last day of each month) of 1-month EDDI and
EDDI constrained by climatology T
air
(EDDI-T),q(EDDI-q), R
d
(EDDI-R
d
), and U(EDDI-U)
for 2011 and 2012. Black box highlights time period shown in (b). (b) Daily time series of
1-month EDDI, EDDI-T, EDDI-q, EDDI-R
d
and EDDI-Ufor May and June 2011 shown to
highlight details of flash drought initiation. Each day in the time series uses the previous 30-day
accumulated E
0
. Note that the yaxis of EDDI is inverted. Light green reference line indicates
start of moderate drought classification (EDDI 50.78).
JUNE 2016 M C E V O Y E T A L . 1771

are presented in Fig. 2. For the Pennsylvania case, EDDI
decays rapidly from neutral to severe drought in 10 days
starting around day 180. At the same time, there is a
rapid spike in SPI to moderate wetness and then a slow
decline toward drought and another rapid decline from
days 206 to 210. Both EDDI and SPI converge on
extreme drought at day 210, while the slower-to-respond
SSI never reaches the extreme drought criterion. This
type of signal, where EDDI shows a rapid change and
SPI and SSI slowly move toward drought, could be used
as a warning signal for potential on-the-ground drought
impacts. For the New York case, both EDDI and SPI
FIG. 7. Evolution of (from top to bottom) the USDM, 1-month EDDI, 1-month ESI, 1-month SSI, and 1-month SPI through the spring
and summer of 2012. USDM data are from (a) 1 May, (b) 5 Jun, (c) 3 Jul, and (d) 31 Jul 2012. (e)–(h) EDDI, (i)–(l) ESI, (m)–(p) SSI, and
(q)–(t) SPI are at 1-month time scales at the end of each month. All drought metrics have been converted to USDM categories according
to Table 1.
1772 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

show two distinct rapid declines, the first at day 140 and
the second at days 156 (SPI) and 157 (EDDI), with
EDDI reaching drought threshold prior to SPI. The
EDDI drought peaks about day 160, SPI at 180, and
SSI around 220. For Wisconsin, EDDI and SPI are
closely correlated during the onset of drought, but SPI
maintains a longer and more severe drought. The Ohio–
Indiana case is similar to the Pennsylvania case in that
EDDI shows the flash drought (starting at day 128) well
in advance of the crash of SPI that starts at day 145. In all
four cases, EDDI is able to detect the flash drought prior
to, or at the same time as, SPI and always ahead of the
SSI signal.
Results illustrated in Figs. 5–8of this paper and in the
companion paper (Part I) highlight two major focal
points of this research: 1) EDDI is a leading indicator of
flash drought conditions and 2) a physically based E
0
is required to capture this signal. This reinforces the
work of Hobbins et al. (2012) and Hobbins (2016), who
concluded that T
air
is not always the dominant driver of
E
0
and that temperature-based parameterizations could
lead to false drying (or wetting) signals when used for
drought-monitoring applications. Our findings (illus-
trated in Fig. 5) also contradict the notion that 2012
should be considered a flash drought case over central
Iowa [in contrast to Mo and Lettenmaier (2015)]: our
results clearly indicate a well-established and persistent
drought signal by both EDDI and SSI, with SPI being
the only indicator to signal a rapid transition from wet to
dry over the period of April–July. Figure 5 illustrates
that the flash drought signal appeared in EDDI starting
in May 2011 and in SPI and SSI starting in August 2011.
d. Extended drought in arid to semiarid regions
In this section, we examine whether EDDI can be
used to characterize historical extended droughts over
the western United States. Droughts in arid to semiarid
regions of the United States are generally slower to
develop than in the central United States, primarily
because of the manner in which water resources are both
naturally and anthropogenically stored. Natural water
storage occurs as winter snowpack at high elevations
that typically reach maximum depth in March or April.
During spring and summer snowmelt, runoff is stored
in reservoirs. Hydrologic drought severity in the west is
strongly linked to reservoir storage and streamflow
(McEvoy et al. 2012;Abatzoglou et al. 2014).
Four extended drought case studies using the USDM,
EDDI, SPI, and SSI are shown in Fig. 9. The first case
focuses on the extreme southwestern drought of 2002
(Figs. 9a,e,i,m), with the USDM mapped on 25 June
2002 (Fig. 9a) and the 6-month EDDI, SPI, and SSI
FIG. 8. Flash drought case studies using daily time series of 1-month EDDI (red line),
1-month SPI (blue line), and 1-month SSI (green line). Note that the yaxis of EDDI is inverted
to clearly visualize drought relative to SPI and SSI. Averaging domains are shown as blue
patches in Fig. 2 and include (from top to bottom) Pennsylvania, New York, Wisconsin, and
Ohio–Indiana. Black boxes highlight the periods of flash drought.
JUNE 2016 M C E V O Y E T A L . 1773

mapped for January–June 2002 (Figs. 9e,i,m). All met-
rics show a similar spatial structure of drought extent,
although EDDI and SPI indicate little to no drought in
Montana. Temperatures were lower than normal over
much of Montana, Wyoming, and the northern portions
of Utah and Colorado and slightly above normal for the
Four Corners region (not shown). This indicates that T
air
was likely driving EDDI negative in Montana; however,
T
air
,q, and Umust have all played a role in driving
EDDI in the positive direction over Utah and Colorado.
The second case focuses on the drought of the 2007
water year (from October 2006 through September
2007; Figs. 9b,f,j,n). The USDM on 2 October 2007 in-
dicates 78% (percent area) of the western United States
FIG. 9. Spatial comparison of drought metrics across the western United States. (a)–(d) USDM on 25 Jun 2002, 2 Oct 2007, 31 Mar 2015,
and 1 Sep 2015; (e)–(h) 6-month EDDI in June 2002, 12-month EDDI in September 2007, 6-month EDDI in March 2015, and 6-month
EDDI in August 2015; (i)–(l) 6-month SSI in June 2002, 12-month SSI in September 2007, 6-month SSI in March 2015, and 6-month SSI in
August 2015; and (m)–(p) 6-month SPI in June 2002, 12-month SPI in September 2007, 6-month SPI in March 2015, and 6-month SPI in
August 2015.
1774 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

in at least D0(Fig. 9b). Figure 9f illustrates the 12-month
EDDI ending in September 2007 and has the strongest
spatial coherence and severity when compared to the
USDM, while SSI and SPI (Figs. 9j,n) underrepresent
the spatial extent shown by USDM and EDDI, partic-
ularly over Nevada, Idaho, and western Montana.
The third case highlights the extraordinary snow
drought that occurred during the winter of 2014/15
(Figs. 9c,g,k,o). At the end of March 2015, the USDM
continued to show D3 and D4 over much of California
and Nevada, but little to no drought over the Wash-
ington Cascades and northern Rockies of Idaho and
Montana (Fig. 9c). The period from October 2014
through March 2015 saw near-normal or even slightly
above normal Prcp over much of the Pacific Northwest,
which is reflected in the 6-month SSI (Fig. 9k) and SPI
(Fig. 9o). However, record warmth during this period
led to extremely freezing levels and much of the Cascade
Range measured snowpacks of less than 25% of normal
by the end of March (not shown). Record warmth and
lack of Prcp led to similar snowpack conditions over the
Sierra Nevada. The 6-month EDDI is the only indicator
to reflect the snow drought conditions in the Cascades
and northern Rockies (Fig. 9g). The fourth case
(Figs. 9d,h,l,p) shows that by the end of August 2015, the
USDM showed Washington 100% covered by D2
(13.36%) and D3 (86.64%), with widespread D3 over
northern Idaho and western Montana (Fig. 9d). Six-
month EDDI (Fig. 9h), SSI (Fig. 9l), and SPI (Fig. 9p) all
agree at this point and show widespread D3 and D4
over the Cascades, northern Idaho, and Montana. The
summer of 2015 was a devastating wildfire season for
California and much of the Pacific Northwest, with close
to 10 million acres burned in the United States (National
Interagency Fire Center; https://www.nifc.gov/fireInfo/
fireInfo_stats_totalFires.html), a result of the snow
drought and record winter warmth followed by a hot and
dry summer. The third and fourth cases (Figs. 9g,h)
demonstrate that EDDI is not only a drought indicator
but that it can also potentially serve as a wildfire risk
indicator. Further research on relationships between
EDDI and wildfire risk will be conducted in future
studies.
The potential usefulness of EDDI to aid in the in-
terpretation of hydroclimatic states at multiple time
scales and over long time periods is assessed in Fig. 10,
which illustrates time series of EDDI averaged over the
northern Sierra Nevada for 1979–2014. As the northern
Sierra Nevada provides much of the water resources to
western Nevada and California, the use of multiple
complementary drought metrics for evaluating short
and extended drought in this region is invaluable. EDDI
at the 2-week and 1-month time scales (Figs. 10a,b)
closely correspond to documented heat waves and
FIG. 10. Area-averaged time series of EDDI over the northern Sierra Nevada from 1979 to
2014 aggregated at (a) 2-week, (b) 1-month, (c) 3-month, (d) 6-month, and (e) 12-month time
scales. Red boxes highlight the four most prominent hydrologic droughts during the
time period.
JUNE 2016 M C E V O Y E T A L . 1775

extreme fire weather in the region (Burt 2007;Trouet
et al. 2009). However, the high frequency of the time
series (Figs. 10a,b) makes it difficult to characterize
hydrologic drought. At longer time scales (Figs. 10c–e)
EDDI clearly identifies all of the major documented
hydrologic droughts over the period from 1979 to 2014
(Seager 2007;Weiss et al. 2009;McEvoy et al. 2012). The
ongoing drought that began in late 2011 clearly stands
out as the most severe and longest-duration event of the
analyzed period. Fast recovery of hydrologic droughts
are also well captured by EDDI at nearly all time scales
when compared to known ‘‘drought buster’’ Prcp events
(Ralph and Dettinger 2012;Dettinger 2013) and wet
periods associated with El Niño (1982/83 and 1997/98)
and La Niña (2010/11).
4. Discussion
Correlations of EDDI to NLDAS-2-forced drought
metrics of SSI and SPI indicate that over much of
CONUS, EDDI spatial distributions are generally sim-
ilar to SPI and SSI. However, over parts of CONUS
weak correlations were found. Comparisons of EDDI to
remotely sensed ESI products also show strong corre-
lations over much of CONUS, with the exceptions of the
northeastern United States during spring and over parts
of the western United States during summer. One rea-
son for weak correlations with ESI over the northeast-
ern United States is largely due to energy-limited land
surface energy-balance conditions over the region,
where ET and E
0
are often positively correlated. The
two main reasons why EDDI showed weak correlations
to other drought metrics are 1) that EDDI often is a
leading indicator and so there is a lag present in the time
series and 2) that EDDI can be strongly positive even
when moisture deficits are not present on the ground
but are rapidly being depleted because of high evapo-
rative demand. It can be difficult to distinguish between
drought early warning and false alarm in EDDI at short
time scales, but a false alarm would only occur if soil
moisture was replenished via an intense precipitation
event. Our analysis highlights the advantage of using
EDDI to monitor potential and actual drought devel-
opment. When EDDI is used in combination with ESI,
actual drought stress may be better understood. A key
strength of EDDI is that it can be effectively used to
provide year-round data, with no limitations during cloudy
days or over snow-covered areas.
For drought monitoring in arid and semiarid regions
of the western United States, EDDI aggregation to
longer time scales (3–12 months) is best suited to cap-
ture the complementary relationship found between
ET and E
0
(Bouchet 1963;Hobbins et al. 2004) and
therefore identify and monitor extended hydrologic
droughts typical of this region. Results illustrate that
in most cases, when Prcp deficits at the 3–12-month
time scales were fairly large, EDDI was strongly
positive. The primary limitation of EDDI for hydro-
logic drought monitoring is that during cold droughts
EDDI may not be able to capture severity because of
the sensitivity to T
air
.
5. Summary and conclusions
This work highlights an application and assessment of
EDDI at multiple time scales and for several hydro-
climates as a companion study to Part I. The methods
and results of Part I are reinforced and a CONUS-wide
evaluation is performed by examining EDDI and indi-
vidual evaporative demand components as they relate
to the dynamic evolution of flash drought over the
central United States, characterization of hydrologic
drought over the western United States, and compari-
son to commonly used drought metrics (USDM, SPI,
SSI, and ESI). The major findings from this work are
summarized as follows:
dEDDI was able to identify droughts over CONUS
consistent with SPI, SSI, and ESI.
dFor flash drought monitoring, EDDI showed potential
development and onset of drought up to 2 months in
advance of the USDM and often led SPI and SSI.
dA unique advantage of EDDI is the ability to de-
compose droughts and test the sensitivity of EDDI to
E
0
drivers. High Uand low qplayed a major role in
initiating the 2011 flash drought case in Iowa. This was
followed by extreme positive T
air
anomalies that drove
much of the EDDI drought signal in 2012 and
exacerbated the depletion of SM.
dTracking drought using submonthly data and multiple
drought indices (Fig. 8) can add value to operational
drought monitoring relative to simply using monthly
data. Flash droughts are dynamic, with large changes
in moisture availability possible over a 1-month period,
even when considering a time series run through a
30-day smoothing filter (i.e., a 1-month drought index
time scale).
dDespite being independent of precipitation, EDDI is
able to capture long-term hydrologic and snow
drought in the western United States.
Despite some limitations, EDDI is shown to provide
useful information on the less understood and docu-
mented dynamical processes associated with drought
evolution and persistence. Results highlighted in this
work illustrate the benefits of assimilating physically
based E
0
estimates and EDDI into operational monitoring
1776 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

products such as the USDM. The additional information
and early warning provided by EDDI could greatly
contribute to a stronger understanding of drought evo-
lution and dynamics, land surface–atmosphere interac-
tions, and, perhaps more importantly, reduce and/or
mitigate future adverse societal effects that have been
associated with past droughts. EDDI could also prove
useful and effective for easy-to-implement operational
early warning for agricultural and fire-weather moni-
toring (Ham et al. 2014) and seasonal forecasting of
drought (McEvoy et al. 2016).
Acknowledgments. The NLDAS-2 data were acquired as
part of the mission of NASA’s Earth Science Division and
archived and distributed by the Goddard Earth Sciences
(GES) Data and Information Services Center (DISC).
Dr. Hobbins was supported from the National Integrated
Drought Information System (NIDIS) and from an Inter-
Agency Agreement between the U.S. Agency for Interna-
tional Development (USAID) and NOAA for support to
the Famine Early Warning Systems Network (FEWS NET;
AID-FFP-P-10-00002/006). Dr. Wood was supported by
a NOAA MAPP Grant (NA11OAR4310142). Dr. McEvoy,
Dr. Huntington, and Mr. Morton were supported by
a DRI Maki Endowment for Enhancing Water Resource
Monitoring in Southern Nevada Grant (6223-640-0969),
a U.S. Bureau of Reclamation Climate Analysis Tools
WaterSMART Program Grant (R11AP81454), a U.S.
Geological Survey and DRI Great Basin Cooperative
Ecosystem Study Unit Collaborative Project on Drought
Monitoring and Fallow Field-Tracking through Cloud
Computing of Landsat, MODIS, and Gridded Climate
Data Archives Grant (G15AC00137), and a U.S. Bureau of
Land Management Grant (L13AC00169). Dr. Anderson
and Dr. Hain were supported by a NASA Applied Sciences
Water Resources Grant (NNX12AK90G).
REFERENCES
Abatzoglou, J. T., 2013: Development of gridded surface meteo-
rological data for ecological applications and modelling. Int.
J. Climatol.,33, 121–131, doi:10.1002/joc.3413.
——, R. Barbero, J. Wolf, and Z. Holden, 2014: Tracking in-
terannual streamflow variability with drought indices in the
U.S. Pacific Northwest. J. Hydrometeor.,15, 1900–1912,
doi:10.1175/JHM-D-13-0167.1.
Allen, R. G., I. A. Walter, R. Elliott, T. Howell, D. Itenfisu, and
M. Jensen, 2005: The ASCE standardized reference evapo-
transpiration equation. Rep. 0-7844-0805-X, 59 pp. [Available
online at http://www.kimberly.uidaho.edu/water/asceewri/
ascestzdetmain2005.pdf.]
Anderson, M. C., J. M. Norman, G. R. Diak, W. P. Kustas, and J. R.
Mecikalski, 1997: A two-source time-integrated model for
estimating surface fluxes using thermal infrared remote sens-
ing. Remote Sens. Environ.,60, 195–216, doi:10.1016/
S0034-4257(96)00215-5.
——, ——, J. R. Mecikalski, J. A. Otkin, and W. P. Kustas, 2007a:
A climatological study of evapotranspiration and moisture
stress across the continental United States based on thermal
remote sensing: 1. Model formulation. J. Geophys. Res.,112,
D10117, doi:10.1029/2006JD007506.
——, ——, ——, ——, and ——, 2007b: A climatological study of
evapotranspiration and moisture stress across the continental
United States based on thermal remote sensing: 2. Surface
moisture climatology. J. Geophys. Res.,112, D11112,
doi:10.1029/2006JD007507.
——, C. Hain, B. Wardlow, A. Pimstein, J. R. Mecikalski, and W. P.
Kustas, 2011: Evaluation of drought indices based on thermal
remote sensing of evapotranspiration over the continental
United States. J. Climate,24, 2025–2044, doi:10.1175/
2010JCLI3812.1.
——,——,J.Otkin,X.Zhan,K.Mo,M.Svoboda,B.Wardlow,
and A. Pimstein, 2013: An intercomparison of drought
indicators based on thermal remote sensing and NLDAS-2
simulations with U.S. Drought Monitor classifica-
tions. J. Hydrometeor.,14, 1035–1056, doi:10.1175/
JHM-D-12-0140.1.
Bouchet, R. J., 1963: Évapotranspiration réelle et potentielle, sig-
nification climatique. IAHS Publ.,62, 134–142.
Burt, C. C., 2007: Extreme Weather: A Guide and Record Book.
W. W. Norton and Co., 320 pp.
Cattiaux, J., and P. Yiou, 2013: U.S. heat waves of spring and
summer 2012 from the flow-analogue perspective [in
‘‘Explaining Extreme Events of 2012 from a Climate Per-
spective’’]. Bull. Amer. Meteor. Soc.,94, S10–S13, doi:10.1175/
BAMS-D-12-00021.1.
Dai, A., 2011: Characteristics and trends in various forms of the
Palmer drought severity index during 1900–2008. J. Geophys.
Res.,116, D12115, doi:10.1029/2010JD015541.
Daly, C., R. P. Neilson, and D. L. Phillips, 1994: A statistical–
topographic model for mapping climatological precipitation
over mountainous terrain. J. Appl. Meteor.,33, 140–158,
doi:10.1175/1520-0450(1994)033,0140:ASTMFM.2.0.CO;2.
Dettinger, M. D., 2013: Atmospheric rivers as drought busters
on the U.S. West Coast. J. Hydrometeor.,14, 1721–1732,
doi:10.1175/JHM-D-13-02.1.
Farahmand, A., and A. AghaKouchak, 2015: A generalized
framework for deriving nonparametric standardized in-
dicators. Adv. Water Resour.,76, 140–145, doi:10.1016/
j.advwatres.2014.11.012.
Guttman, N. B., 1999: Accepting the Standardized Precipitation
Index: A calculation algorithm. J. Amer. Water Resour. Assoc.,
35, 311–322, doi:10.1111/j.1752-1688.1999.tb03592.x.
Ham, C., M. T. Hobbins, K. L. Abt, and J. P. Prestemon, 2014:
Using the Evaporative Demand Drought Index and the
Palmer drought severity index to forecast the number of large
wildland fires on federal lands. Large Wildland Fires Conf.,
Missoula, MT, Association for Fire Ecology and the In-
ternational Association of Wildland Fire. [Available online
at http://largefireconference.org/wp-content/uploads/2013/06/
Oral-Presentation-Abstracts-V4.pdf.]
Han, S., F. Tian, and H. Hu, 2014: Positive or negative cor-
relation between actual and potential evaporation? Eval-
uating using a nonlinear complementary relationship
model. Water Resour. Res.,50, 1322–1336, doi:10.1002/
2013WR014151.
Hao, Z., and A. AghaKouchak, 2014: A nonparametric multivariate
multi-index drought monitoring framework. J. Hydrometeor.,
15, 89–101, doi:10.1175/JHM-D-12-0160.1.
JUNE 2016 M C E V O Y E T A L . 1777

Hargreaves, G. H., and Z. A. Samani, 1985: Reference crop
evapotranspiration from temperature. Appl. Eng. Agric.,1,
96–99, doi:10.13031/2013.26773.
Heddinghaus, T. R., and P. Sabol, 1991: A review of the Palmer
drought severity index and where do we go from here? Pre-
prints, Seventh Conf. on Applied Climatology, Boston, MA,
Amer. Meteor. Soc., 242–246.
Hobbins, M. T., 2016: The variability of ASCE standardized
reference evapotranspiration: A rigorous, CONUS-wide
decomposition and attribution. Trans. ASABE,59, 561–576,
doi:10.13031/trans.59.10975.
——, J. A. Ramírez, and T. C. Brown, 2004: Trends in pan evap-
oration and actual evaporation across the conterminous U.S.:
Paradoxical or complementary? Geophys. Res. Lett.,31,
L13503, doi:10.1029/2004GL019846.
——, A. Dai, M. L. Roderick, and G. D. Farquhar, 2008: Revisiting
the parameterization of potential evaporation as a driver of
long-term water balance trends. Geophys. Res. Lett.,35,
L12403, doi:10.1029/2008GL033840.
——, A. Wood, D. Streubel, and K. Werner, 2012: What drives the
variability of evaporative demand across the conterminous
United States? J. Hydrometeor.,13, 1195–1214, doi:10.1175/
JHM-D-11-0101.1.
——, ——, D. J. McEvoy, J. L. Huntington, C. Morton, and
J. Verdin, 2016: The Evaporative Demand Drought Index.
Part I: Linking drought evolution to variations in evapora-
tive demand. J. Hydrometeor.,17, 1745–1761, doi:10.1175/
JHM-D-15-0121.1.
Jensen, M. E., 1973: Consumptive Use of Water and Irrigation
Water Requirements. American Society of Civil Engineers,
215 pp.
Karl, T. R., and Coauthors, 2012: U.S. temperature and drought:
Recent anomalies and trends. Eos, Trans. Amer. Geophys.
Union,93, 473–474, doi:10.1029/2012EO470001.
Koster, R. D., and Coauthors, 2004: Regions of strong coupling
between soil moisture and precipitation. Science,305, 1138–
1140, doi:10.1126/science.1100217.
——, and Coauthors, 2006: GLACE: The Global Land–
Atmosphere Coupling Experiment. Part I: Overview.
J. Hydrometeor.,7, 590–610, doi:10.1175/JHM510.1.
——, S. D. Schubert, and M. J. Suarez, 2009: Analyzing the con-
currence of meteorological droughts and warm periods, with
implications for the determination of evaporative regime.
J. Climate,22, 3331–3341, doi:10.1175/2008JCLI2718.1.
Liang, X., D. P. Lettenmaier, E. F. Wood, and S. J. Burges, 1994: A
simple hydrologically based model of land surface water and
energy fluxes for general circulation models. J. Geophys. Res.,
99, 14 415–14 428, doi:10.1029/94JD00483.
McEvoy, D. J., J. L. Huntington, J. T. Abatzoglou, and L. M.
Edwards, 2012: An evaluation of multiscalar drought indices
in Nevada and eastern California. Earth Interact.,16,
doi:10.1175/2012EI000447.1.
——, ——, J. F. Mejia, and M. T. Hobbins, 2016: Improved sea-
sonal drought forecasts using reference evapotranspiration
anomalies. Geophys. Res. Lett.,43, 377–385, doi:10.1002/
2015GL067009.
McKee, T. B., N. J. Doesken, and J. Kleist, 1993: The relationship
of drought frequency and duration to time scales. Preprints,
Eighth Conf. on Applied Climatology, Anaheim, CA, Amer.
Meteor. Soc., 179–184.
Mesinger, F., and Coauthors, 2006: North American Regional
Reanalysis. Bull. Amer. Meteor. Soc.,87, 343–360, doi:10.1175/
BAMS-87-3-343.
Milly, P. C. D., and K. A. Dunne, 2011: On the hydrologic ad-
justment of climate-model projections: The potential pitfall of
potential evapotranspiration. Earth Interact.,15, doi:10.1175/
2010EI363.1.
Mo, K. C., and D. P. Lettenmaier, 2015: Heat wave flash droughts
in decline. Geophys. Res. Lett.,42, 2823–2829, doi:10.1002/
2015GL064018.
Monteith, J. L., 1965: Evaporation and environment. Symp. Soc.
Exp. Biol.,19, 205–234.
Mu, Q., M. Zhao, J. S. Kimball, N. G. McDowell, and S. W.
Running, 2013: A remotely sensed global terrestrial drought
severity index. Bull. Amer. Meteor. Soc.,94, 83–98,
doi:10.1175/BAMS-D-11-00213.1.
Otkin, J. A., M. C. Anderson, C. Hain, I. E. Mladenova, J. B.
Basara, and M. Svoboda, 2013: Examining rapid onset drought
development using the thermal infrared–based evaporative
stress index. J. Hydrometeor.,14, 1057–1074, doi:10.1175/
JHM-D-12-0144.1.
——, ——, ——, and M. Svoboda, 2014: Examining the relation-
ship between drought development and rapid changes in the
evaporative stress index. J. Hydrometeor.,15, 938–956,
doi:10.1175/JHM-D-13-0110.1.
Palmer, W. C., 1965: Meteorological drought. U.S. Weather
Bureau Research Paper 45, 58 pp. [Available online
at http://www.ncdc.noaa.gov/temp-and-precip/drought/docs/
palmer.pdf.]
Priestley, C. H. B., and R. J. Taylor, 1972: On the assessment
of surface heat flux and evaporation using large-scale pa-
rameters. Mon. Wea. Rev.,100, 81–92, doi:10.1175/
1520-0493(1972)100,0081:OTAOSH.2.3.CO;2.
Quiring, S. M., 2009: Developing objective operational definitions
for monitoring drought. J. Appl. Meteor. Climatol.,48, 1217–
1229, doi:10.1175/2009JAMC2088.1.
Ralph, F. M., and M. D. Dettinger, 2012: Historical and national
perspectives on extreme West Coast precipitation associated
with atmospheric rivers during December 2010. Bull. Amer.
Meteor. Soc.,93, 783–790, doi:10.1175/BAMS-D-11-00188.1.
Roderick, M. L., M. T. Hobbins, and G. D. Farquhar, 2009: Pan
evaporation trends and the terrestrial water balance. II. En-
ergy balance and interpretation. Geogr. Compass,3, 761–780,
doi:10.1111/j.1749-8198.2008.00214.x.
Seager, R., 2007: The turn of the century North American drought:
Global context, dynamics, and past analogs. J. Climate,20,
5527–5552, doi:10.1175/2007JCLI1529.1.
——, M. Hoerling, S. Schubert, H. Wang, B. Lyon, A. Kumar,
J. Nakamura, and N. Henderson, 2015: Causes of the 2011–14
California drought. J. Climate,28, 6997–7024, doi:10.1175/
JCLI-D-14-00860.1.
Sheffield, J., E. F. Wood, and M. L. Roderick, 2012: Little change in
global drought over the past 60 years. Nature,491, 435–438,
doi:10.1038/nature11575.
Svoboda, M., and Coauthors, 2002: The Drought Monitor.
Bull. Amer. Meteor. Soc.,83, 1181–1190, doi:10.1175/
1520-0477(2002)083,1181:TDM.2.3.CO;2.
Thornthwaite, C. W., 1948: An approach toward a rational classi-
fication of climate. Geogr. Rev.,38, 55–94, doi:10.2307/210739.
Trouet, V., A. H. Taylor, A. M. Carleton, and C. N. Skinner, 2009:
Interannual variations in fire weather, fire extent, and
synoptic-scale circulation patterns in Northern California and
Oregon. Theor. Appl. Climatol.,95, 349–360, doi:10.1007/
s00704-008-0012-x.
van der Schrier, G., P. D. Jones, and K. R. Briffa, 2011: The sensi-
tivity of the PDSI to the Thornthwaite and Penman–Monteith
1778 JOURNAL OF HYDROMETEOROLOGY VOLUME 17

parameterizations forpotential evapotranspiration.J. Geophys.
Res.,116, D03106, doi:10.1029/2010JD015001.
Vicente-Serrano, S. M., J. I. López-Moreno,S. Beguería, J. Lorenzo-
Lacruz, C. Azorin-Molina, and E. Morán-Tejeda, 2012: Accu-
rate computation of a streamflow drought index. J. Hydrol.
Eng.,17,318–332,doi:10.1061/(ASCE)HE.1943-5584.0000433.
Weiss, J. L., C. L. Castro, and J. T. Overpeck, 2009: Distinguishing
pronounced droughts in the southwestern United States:
Seasonality and effects of warmer temperatures. J. Climate,
22, 5918–5932, doi:10.1175/2009JCLI2905.1.
Xia, Y., and Coauthors, 2012a: Continental-scale water and energy
flux analysis and validation for the North American Land
Data Assimilation System project phase 2 (NLDAS-2):
1. Intercomparison and application of model products.
J. Geophys. Res.,117, D03109, doi:10.1029/2011JD016048.
——, and Coauthors, 2012b: Continental-scale water and energy
flux analysis and validation for the North American Land
Data Assimilation System project phase 2 (NLDAS-2):
2. Validation of model-simulated streamflow. J. Geophys. Res.,
117, D03110, doi:10.1029/2011JD016051.
Yao, Y., S. Liang, Q. Qin, and K. Wang, 2010: Monitoring drought
over the conterminous United States using MODIS and
NCEP Reanalysis-2 data. J. Appl. Meteor. Climatol,49, 1665–
1680, doi:10.1175/2010JAMC2328.1.
JUNE 2016 M C E V O Y E T A L . 1779