Content uploaded by Zhanjiang Sha
Author content
All content in this area was uploaded by Zhanjiang Sha on Feb 27, 2014
Content may be subject to copyright.
Online system, http://www.scar.ac.cn
Sciences in Cold and Arid Regions
2009, 1(5): 0467–0474
Using remote-sensing technologies in combination with
Cesium-137 measurements to estimate soil-erosion
quantity in semi-arid steppe areas
ZhanJiang Sha 1, 2*, HaiZhou Ma 1, LingQin Li 2 , Jinzhou Du 3,
FeiQuan Wu 1,4 , QiShun Fan 1, 4
1. Qinghai Institute of Salt Lakes, Chinese Academy of Sciences, Xining, Qinghai 810008, China
2. Department of Geography, Qinghai Normal University, Xining, Qinghai 810008, China
3. State key laboratory of estuarine and coastal research, Shanghai, 200062, China
4. Graduate University of Chinese Academy Sciences, Beijing 100049, China
*Correspondence to: ZhanJiang Sha, Ph.D., of Qinghai Institute of Salt Lakes, Chinese Academy of Sciences, Xining, Qinghai
810008, China. Email: sazhanjiang@sina.com
Received: 20 January 2009 Accepted: 3 June 2009
ABSTRACT
Soil erosion by wind is one of the most important processes in the changing the earth’s surface in semi-arid areas, Thus it is of
great importance to study soil-erosion action. Using integrated technologies of remote sensing and geochemistry radioactivity iso-
tope to extract regional soil-erosion information and to calculate quantity of soil erosion is accomplished successfully in this paper
by means of beneficial experiments in the Talatan region of the Gonghe Basin, which is located in northeastern Qinghai-Tibet Pla-
teau in China. The results show that the soil erosion by wind is not intensive in this region; the erosion types belong to the classes
of very-soft erosion and soft-erosion type, which account for 47.12 percent and 35.58 percent, respectively, of the total study area.
In total, two kinds of soil erosion account for 82.70 percent of the study area; only a small area belongs to the classes of severe
erosion and very-severe erosion; this area is about 22.14 km2. Severe deposition activity has taken place in this region, and has
appeared in a large area (322.67 km2), which accounts for 11.78 percent of the total study area. The results of this study show that
soil erosion and deposition inventories are 870,000–1,150,000 tons and 550,000–780,000 tons, respectively, per year. The soil in-
ventory shows about 320,000–370,000 tons from Talatan to Longyangxia reservoir per year. Using remote-sensing technology and
137Cs techniques is a valid means to analyze and to evaluate the quantity of soil erosion by wind in semi-arid environments.
Keywords: remote sensing; cesium-137; soil erosion; wind erosion; China
1. Introduction
Remote-sensing technologies have many useful attrib-
utes to provide plentiful information in the study of envi-
ronment. For instance, with remote-sensing macro and
multi-platform information about the earth’s surface can be
obtained timely for quick updating data, and especially
monitoring regional environment changes. The application
of remote-sensing technologies is a very useful means for
studying soil erosion. The remote-sensing models of soil
erosion by wind have been being developed since the 1950s.
These models primarily include the wind erosion equation
(WEQ) (Woodruff and Siddoway, 1965) and the revised
wind erosion equation (RWEQ) (Fryrear et al., 1994), the
wind erosion appraisal model (WEAM) (Shao et al., 1996),
wind erosion prediction system (WEPS) (Hagen, 2004), etc.
These models may be used to forecast regional soil ero-
sion of cropland and grassland. However, because many
kinds of models cannot be applied to large areas to quantita-
tively calculate soil erosion by wind, it is still difficult to
evaluate regional soil erosion. Study of regional soil erosion
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
468
in large grassland area needs to be done in combination with
other measurements. Thus, in this paper, we used remote-
sensing techniques in combination with Cesium-137 (137Cs)
measurements to estimate the quantity of soil loss by erosion.
Many hydrologic studies have used measurements of the
radiogenic isotope 137Cs to calculate quantity of soil erosion.
137Cs has been used in experiments of atomic energy or nu-
clear leaking from the 1950s through the 1970s. 137Cs is
absorbed intensively by emplastic minerals and organisms in
soil layers. It clings with soil grain so tightly that it cannot be
dissolved or moved singly; if 137Cs is moved on land surface
by either water or wind, it is accompanied with soil grains to
erode and deposit (Walling and He, 1999). Thus, soil erosion
and deposition define the movement and redistribution of
137Cs in soil layers; 137Cs can also be used to follow the route
and movement of soil erosion (Zhang et al., 1990; Soto and
Navas, 2004; Visser et al., 2005).
Many researchers have studied water erosion by using
137Cs measurement and have made great progress on accu-
rately determining the rate of water erosion and redistribu-
tion of soil material in small watershed; at the same time, the
well-considered study theory and simulation models have
been produced (Schuller et al., 2004).
On the aspects of a wind erosion study, however, there
are many difficulties in using 137Cs measurement because
there is a great deal of randomness in the course of wind
erosion compared with that of a water-erosion rate. Thus, the
study of wind erosion with 137Cs measurements is weak in
comparison. But in recent years, there have been many use-
ful studies employing 137Cs measurements to accurately
determine the rate of soil erosion by wind (Collins et al.,
2001; Walling and He, 2003; Li et al., 2005), which have
provided some positive results.
The Talatan area of the Gonghe Basin in China’s Qing-
hai Province has been used as a study area for this paper.
Our intention, in this paper, is to try to calculate the quantity
of soil erosion of this region using remote-sensing technolo-
gies combined with 137Cs measurement and to try to explore
the quantitative methods of wind erosion of soil in the re-
gions under the same environmental conditions.
2. Study area
Talatan, which covers about 2,750 km2, is located in the
Gonghe basin of Qinghai Province (Figure 1). It is between
35º45'–36º20'N, 100º00'–101º00'E, and has elevations rang-
ing from 2,600–3,300 m. The annual average temperature is
0.86 ºC, the annual average amount of precipitation is 280
mm, the evaporation capacity is 1,620 mm, and the region
belongs to the high semi-arid steppe climate type. Topogra-
phy of Talatan is mostly flat with a slightly wavy terrain
except the surrounded hilly areas and the incised river valley
in the region. In the lower elevations, the landscape becomes
a desert steppe, made up of flow dune, semi-flow dune,
semi-fixed dune, fixed dune, and different types of grassland.
The soil of land’s surface is mainly windborne sand and
loess, sandy paloesoil, and deposited substances of rivers
and lakes; the soil profile is made of windborne sand, loess,
and paloesoil.
Figure 1 Location of Talatan (study area) in the northeastern of Qinghai Province, China. The satellite image shows the Talatan area of the
Gonghe Basin in 2001.
Three subregions including the windborne-erosion area,
the flow-dune deposition area, and the loess-deposition area,
are distributed in the northwestern and southeastern portions
of Talatan and at the feet of the mountain edges. Vegetation
is primarily made up of cogongrass and sedge, etc. The
cover rate of vegetation is very low in the region, ranging
from about 10 percent to 90 percent; but for most of this
study region, that cover rate is less than 30 percent, and de-
sertification of land has taken place widely. Hence, the eco-
logical environment of Talatan is very fragile. For all intents
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
469
and purposes, the environmental characteristics of Talatan are
arid, low precipitation, windy, and with little surface water.
This is a typical region that has soil erosion by wind.
3. Methods
3.1. Obtaining soil-erosion information using re-
mote-sensing technologies
The classification of soil erosion by wind was defined
according to the environmental characteristics of the region,
which were determined based on field investigation. The soil
erosion was classified into five types: very soft erosion, soft
erosion, medium erosion, severe erosion, and very severe
erosion. The classification of soil erosion by wind has obvious
local characteristics, which is a relative notion, but the classi-
fication criterion may be detailed and accurate to explain how
soil erosion by wind is a different condition in the region.
The key environmental factors to influence soil erosion
by wind were extracted from the remote-sensing information
we gathered; these factors included the terrain and its fea-
tures, surface materials, and the cover rate of vegetation. For
this reason, we obtained the soil-erosion data from re-
mote-sensing information based primarily on three aspects:
the cover rate of vegetation, the information of land deserti-
fication, and land use in practical work.
The data of p133r035 Enhanced Thematic Mapper
(ETM) (July 12, 2001) was used as the primary information
from which to extract key environmental factors. The ETM
data was fused, corrected, and classified to obtain existing
thematic information of soil erosion in ERDAS IMAGINE.
These results were then modified in ArcGIS referring to the
data of geology, physiognomy, soil, vegetation, hydrology,
and climate of the region, then the thematic information of
environmental factors were obtained, which included the
data of vegetation cover rate, land desertification, and land
use (Figure 2a, b, and c). The data of slope was calculated
using a 30-m resolution Digital Elevation Model (DEM) in
hilly areas and in the valley of the study area.
Figure 2 Thematic information maps related to key environment factors effecting soil erosion and the class of soil erosion: (a) cover rate of
vegetation in study area; (b) land use of study area; (c) desertified land of study area; (d) class of soil erosion in study area.
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
470
All of these data were analyzed and calculated in Ar-
cGIS, This included spatial overlay, Boolean operating,
small polygon sieve, attribute merge and recode, etc, and the
spatial distributing information of soil erosion by wind was
obtained (Figure 2d). The data of soil erosion was evaluated
in class precision with ERDAS 8.6 software (ERDAS, Inc.
1997) and was calculated for the area of all types of soil
erosion in ArcGIS 9.0 software (ESRI, Inc. 1996).
3.2. Collection and determination of 137Cs samples
Before collecting the 137Cs soil samples in the field, we
determined the positions of the sampling spots in the labora-
tory, according to the map of soil-erosion type that had been
obtained using remote-sensing technologies; the positions
may be adjusted according to field working. Three to six sam-
ples of each type of soil erosion were collected (Figure 3).
To collect, at the same time, separate soil samples and
mixed soil samples of 137Cs in each collection spot, the sam-
ples were collected by volume–weight. We used a round
sampling container, with a radius of 7.5 cm, to contain sepa-
rate soil samples. The samples were obtained at 5 cm inter-
vals from land surface to bottom vertically. The total depth
of sampling was 30 cm. In total, we have collected six sepa-
rate soil samples of 137Cs and six volume-weight soil sam-
ples. We used a round sampling container, with a radius of
7.5 cm and length of 15 cm, to contain mixed soil samples.
The mixed soil samples of 137Cs were collected to a total
depth of 30 cm from land surface to the bottom at the sam-
pling spot. The interval was 5 cm. The sampling date for all
was from May 17 through May 22, 2005.
Because 137Cs is absorbed primarily on the surface of
fine mineral clay, the large-grain mineral was very difficult
to absorb 137Cs, and its quantity of adsorption was ignored in
calculating; the fine grain of soil was selected and measured
(
φ
≤ 2.0 mm). The soil samples were dried, ground, sieved
(
φ
= 2.0 mm), and weighed to measure the 137Cs concentra-
tion of the soil.
The soil samples of 137Cs were measured by nucleus
laboratory of the physics department of Lanzhou University.
The machine used to detect 137Cs concentration of soil is a
high pure Cr probe multi-ways γ apparatus of ORTEC
Company of the United States. The sensitive volume of
probe is 100 cm3, and efficiency of detection is 4 percent;
500 g of soil samples were needed, the time needed for de-
tection was about 21,600 s to 54,000 s, and the relative error
of a repeat measurement was less than 10 percent.
Figure 3 Sites of sampling spots in the study area
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
471
4. Result and analysis
4.1. Determining the reference value of 137Cs in the study area
Another researcher had already measured the reference
value of 137Cs in the study area, and the results showed that it
is 2319 Bq/m2 in Gonghe Basin (Zhang et al., 2002), and it
is 2376.04 Bq/m2 in northern area of Qinghai-Tibet Plateau
(Yan et al., 2000). The reference value of 137Cs was deter-
mined for this paper according to the study result of Zhang
et al.; thus, the reference value is equal to 2319 Bq/m2.
4.2. The soil-erosion model
The 137Cs concentration has difference in the soil profile
according to the measurement result of each sampling soil
profile, but the depth of 137Cs is about 30 cm from land sur-
face to bottom. The 137Cs concentration is distributed ac-
cording to exponent function in the soil profile, which re-
duces gradually from the top to the bottom of the profile and
80% of 137Cs concentration is distributed in 10 cm from the
top to the bottom. The profile-distribution model of soil 137Cs
concentration is Cs = ae-bz (a, b > 0) (Tang et al., 2000) ac-
cording to analysis of the 137Cs data of each soil-sampling
profile. Using our measured data to build the model of 137Cs
concentration in soil profile, the parameters a and b were
calculated as a = 40.02, b = 0.023, so the model of the study
area is
)0(02.40 023.0 HzeCs Z≤≤= − (1)
where z is the depth of soil; and H is maximum depth of
137Cs distribution in the soil layer.
The quantity of soil erosion was calculated by functions
)0,( >= −baaeCs bZ
(2)
tref CsCsCs −=Δ
thus,
dzaeDCsCsCs HbZ
tref ∫−
⋅=−=Δ 0
∫∫ −− ⋅=⋅− hbZ
H
h
bZ dzaeDdzaeD 0 (3)
The total soil-erosion thickness (h) in t years was calcu-
lated by
(
)
⎥
⎦
⎤
⎢
⎣
⎡
⋅
⋅−
−⋅−= Da
bCsCs
b
htref
1ln
1 (4)
Putting a and b in function (4), we obtained
(
)
⎥
⎦
⎤
⎢
⎣
⎡
⋅
−⋅−
−⋅
−
−= D
CsCs
htref
02.40
)023.0(
1ln
023.0
1(5)
where D is the soil volume-weight (g/cm3); Csref is reference
inventory of 137Cs in the study area (Bq/m2); Cst is sampling
inventory of 137Cs (Bq/m2).
Using Lowance model (Lowance et al., 1988) to cal-
culate soil-depositing quantity was fairly accurate, as
shown by many study results. The total depositing
thickness of soil (h, unit: mm) in t years was calculated
by
w
reft
CsD
CsCs
h⋅
−
= (6)
where, D is the soil volume-weight (g/cm3); Csw is the
average 137Cs concentration of soil profile (Bq/kg).
From all of the preceding conclusions, the soil ero-
sion or deposition quantity of annual average was calcu-
lated in each sampling spot according to the function
1000×
×
=
rr hDE (7)
where Er is mean annual soil erosion or deposition quan-
tity (t/km2·yr); hr = h/t, t = T−1964, T is the year of sam-
pling. The maximum deposition of 137Cs in soil was in
year 1964 (Table 1).
4.3. Analyzing soil erosion quantity integrated with data of
remote sensing and 137Cs
Based on the spatial distribution information of soil ero-
sion by remote-sensing technologies and combined with
137Cs data of soil erosion, the quantity of soil erosion in the
study area was estimated (Table 2 and Table 3).
In Table 1, several pieces of data need to be explained.
Firstly, the value of sample QTL021 and QTL022 are much
lower than the regional reference value; the erosion quantity
is 126.56 and 189.84 t/km2·yr, because the sample were col-
lected at the foot of the hill where the grassland was main-
tained well by human beings, and the cover rate of vegeta-
tion is greater than 70 percent, so the rate of soil erosion is
much lower than the quantity of the soft-erosion class.
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
472
Table 1 The
137Cs inventory and the rate of soil erosion in study area
Samples No. V-weight
(g/cm3)
137Cs inventory
(Bq/m2)
Rate of erosion
（t/km2·yr） Environmental features of sampling spots
QTL001 1.37 1,224.78 892.58 brush-dune, plat, cover rate of vegetation 10%
QTL002 1.24 989.52 1,259.19 brush-dune, plat, cover rate of vegetation 10%
QTL003 1.36 1,326.00 784.68 arid-grassland, plat, cover rate 30%
QTL004 1.25 795.00 1,598.99 arid-grassland, river terrace, cover rate 20%
QTL005 1.50 1,152.00 942.67 arid grassland, plat, cover rate 30%
QTL006 1.34 755.76 1,577.38 arid-grassland, shallow hole, cover rate 10%
QTL007 1.27 982.98 1,249.57 river terrace, loess, cover rate <10%
QTL008 1.66 821.70 1,286.12 river terrace, loess, cover rate <10%
QTL009 1.62 1,632.96 479.12 river terrace, brush-dune, cover rate 20%
QTL010 1.48 1,523.23 580.13 arid-grassland, flat, cover rate 30%
QTL011 1.67 1,042.08 1,025.05 arid-grassland, river terrace, cover rate 20%
QTL012 1.19 2,081.31 153.87 grassland, flat, cover rate >70%
QTL013 1.49 1,832.70 220.28 brush-dune, plat, cover rate 20%
QTL014 1.52 1,896.96 280.22 grassland, flat, cover rate 50%
QTL015 1.31 10,650.30 −7,498.24 dry lakebed, cover rate <10%
QTL016 1.62 2,703.96 −1,688.72 dune foot, cover rate <10%
QTL017 1.38 2,326.68 −33.33 flat sand land, cover rate 20%
QTL018 1.24 10,602.00 −7,088.58 dry lakebed, cover rate <10%
QTL019 1.72 2,198.16 75.17 flat sand land, cover rate 10%
QTL020 1.57 2,929.62 −2,394.40 dune foot, cover rate <10%
QTL021 1.15 2,121.75 126.56 mountain grassland foot, cover rate 90%
QTL022 1.30 2,028.00 189.84 arid-grassland, flat, cover rate 70%
QTL023 1.22 1,822.68 344.55 arid-grassland, river terrace, cover rate 50%
QTL024 1.56 1,291.68 786.74 arid-grassland, flat, cover rate 50%
Secondly, the values of samples QTL017 and QTL019
are also very low. The rates of soil erosion or deposition in
these sampling spots are small, because these samples
were collected in the wide, flat area of sand land where the
erosion and deposition come to a balanced condition. Thus
the values of these samples do not reflect deposition of
sand in this area.
Finally, the samples QTL015 and QTL018 that were
collected nearby the dry lakes (Lake of Dalian and Lake of
Genga) need to be explained. The two lakes have accom-
panied the changing of climate and alternated in dry and
wet conditions in recent years (Dong et al., 1993). The
lakes were dry in year of sampling (May 2005) when soil
samples were collected. The 137Cs concentrations of two
samples are very high and imply that deposition is the
main action in these spots. It is logical to explain that the
soil grains carried by wind or water were obstructed to
deposit in lakes, and that the deposition was the main ac-
tion for long time; however, in recent years, the lakebeds
were eroded when they were exposed without water.
Except for those abnormal data, the corresponding re-
lationships were build up between the class of soil erosion
and the quantity of soil erosion using remote-sensing (RS)
techniques and 137Cs (Table 2). The erosion quantity of
each class was calculated by using the area of each soil
erosion class from RS data in ArcGIS 9.0 version software
and multiplying it with the corresponding quantity of av-
erage annual soil erosion. Adding to the erosion quantity of
each class, the inventory of erosion and deposition was
measured in this study area (not including the deposition
quantity in the lakes with water) (Table 3). The results
show that soil erosion by wind is not intense in this region;
the results indicate that these classes of erosion belong to
the very-soft erosion and the soft-erosion types, which
account for 47.12 percent and 35.58 percent, respectively,
of the total study area; in total the two types of soil erosion
account for 82.70 percent of the study area; the erosion
velocity is 220.28–580.13 t/(km2·yr), and only a small area
belongs to the severe-erosion and very-severe erosion
classes, it is about 22.14 km2. If severe depositing action
has taken place in the region, and it appears that the large
area with sand deposition, which is 322.67 km2, accounts
for 11.78% of studied total area. The result shows that soil
erosion and deposition inventory is 870,000–1,150,000
tons and 550,000–780,000 tons, respectively, per year as
shown by calculating the soil-erosion inventory. The soil
inventory is assuming 320,000–370,000 tons from Talatan
to Longyangxia Reservoir per year.
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
473
Table 2 Corresponding relationship of soil-erosion class and soil-erosion quantity
Class of soil erosion The rate of erosion（t/km2·yr）
Very soft erosion 220.28–344.55
Soft erosion 479.12–580.13
Medium erosion 784.68–892.58
Severe erosion 942.67–1,025.05
Very severe erosion 1,249.57–1,598.99
Dune (severe deposition) 1,688.72–2,394.40
Dry lakebed (very severe deposition) 7,088.58–7,498.24
Table 3 Quantity of soil erosion of study area
Class of soil erosion Area (Km2) Quantity of soil erosion（t/yr）
Very soft erosion 1,290.3016 284,227.63–444,573.42
Soft erosion 974.2607 466,787.79–565,197.86
Medium erosion 128.9444 101,180.09–115,093.19
Severe erosion 21.5147 20,281.26–22,053.64
Very severe erosion 0.6259 782.11–1,000.81
Dune (severe deposition) 321.4833 542,895.28–769,759.61
Dry lakebed (very severe deposition) 1.1818 8,377.28–8,861.46
Inventory of erosion 873,258.88–1,147,919.59
Inventory of deposition 551,272.56–778,621.07
Inventory entering reservoir 321,986.32–369,298.52
5. Conclusions
From the study presented here, it can be concluded that
satellite images, if employed judiciously, can be a very ef-
fective tool for mapping soil erosion. This is especially true
in hilly arid terrain like the Talantan area of Gonghe Basin
on the Qinghai-Tibet Plateau, where poor accessibility and
dense vegetation cover hinder fieldwork; in these conditions,
images from satellite technology can play a significant role
in obtaining the data of key environment factors that affect
soil erosion—such as land use, desertification of land, and
the cover rate of vegetation. Using remote-sensing technolo-
gies to obtain the information of soil erosion is one of the
valid means in the study area. The class of soil erosion by
wind was defined according to environmental characteristics
of the region, based on field investigation. The soil erosion
was divided into five classifications such as very soft erosion,
soft erosion, medium erosion, severe erosion, and very se-
vere erosion; the dune land and dry lakebed were classified
in severe erosion and very severe erosion, respectively.
Whereas the data of soil erosion were obtained with ETM
(Enhanced Thematic Mapper) data in ERDAS and ArcGIS,
the quantity of soil erosion was not obtained with re-
mote-sensing data but was measured by 137Cs techniques.
These data included the area of each erosion class and spa-
tial distribution map.
It is accurate and valid to investigate the quantity of re-
gional soil erosion in this study area and calculate the
soil-loss quantity of each sampling spots through the appli-
cation of 137Cs techniques, and thus can provide useful data
for studying soil erosion, but it cannot effectively measure
soil-loss quantity over large areas. With the aid of re-
mote-sensing data, a fieldwork program can be planned in
advance based on prior knowledge so that the sites for col-
lecting 137Cs soil samples are validly determined. With the
data of soil erosion created by remote-sensing technologies,
the inventory of soil erosion and deposition are calculated
quickly and successfully. Although there are many questions
that need to further study during the application of re-
mote-sensing technologies associated with 137Cs techniques,
it is of great significance to get positive results in this area by
making full use of them to measure the quantity of regional
soil erosion and to quantitatively study soil erosion.
Acknowledgments:
The 137Cs soil samples were processed by the Department of
Physics of Lanzhou University. The paper is financed by the
Institute of Qinghai Salt Lakes, Chinese Academy of Sci-
ences (contract KZCX2-SW-118). The open fund of State
Key Laboratory of Estuarine and Coastal Research also
supported this research. We thank Lu HuaYu and Li Xiao-
Qiang (the institute of Earth Environment, CAS) for their
valuable comments to an earlier draft of this paper. The
quality of the article was greatly enhanced by the sugges-
tions made by four anonymous reviewers.
Zhanjiang Sha et al., 2009 / Sciences in Cold and Arid Regions, 1(5): 0467–0474
474
REFERENCES
Collins AL, Walling DE, Sichingabula HM, Leeks GJL, 2001. Using 137Cs
measurements to quantify soil erosion and redistribution rates for areas
under different land uses in the Upper Kaleya River basin, southern
Zambia. Geoderma, 104: 299–323.
Dong GR, Gao SY, Jin J, 1993. Study of desertification and prevention
approach in Gonghe Basin of Qinghai Province. Science Publishing
Company, China.
Fryrear DW, Saleh A, Bilbro JD, Zobeck TM, Stout JE, 1994.
Field-tested wind-erosion model. In: Buerkert B, Allision BE, von
Oppen M, Verlag M. Proc. of International Symposium "Wind Ero-
sion in West Africa: The Problem and Its Control." Weikersheim,
Germany.
Hagen LJ, 2004. Evaluation of the wind-erosion prediction system
(WEPS) erosion submodel on cropland fields. Environmental Mod-
elling and Software, 19(2): 171–176.
Li M, Li ZB, Liu PL, Yao WY, 2005. Using Cesium-137 technique to study
the characteristics of different aspects of soil erosion in the Wind/water
Erosion Crisscross Region on Loess Plateau of China. Applied Radiation
and Isotopes, 62: 109–113.
Lowance R, Mclntyre S, Lance C, 1988. Erosion and deposition in a
field/forest system estimated using 137Cs activity. Journal of Soil and Wa-
ter Conservation, 43(2): 195–198.
Schuller P, Walling DE, Sepulveda A, Trumper RE, Rouanet JL, Pino I,
Castillo A, 2004. Use of 137Cs measurements to estimate changes in
soil-erosion rates associated with changes in soil-management practices
on cultivated land. Applied Radiation and Isotopes, 60: 759–766.
Shao Y, Raupch MR, Leys JF, 1996. A model for predicting Aeolian
sand drift and dust entrainment on scales from paddock to region.
Australian Journal of Soil Research, 34: 309–342.
Soto J, Navas A, 2004. A model of 137Cs activity profile for soil-erosion
studies in uncultivated soils of Mediterranean environments. Journal of
Arid Environments, 59: 719–730.
Tang XY, Yang H, Zhao QG, 2000. Review of Progress in Application of
137Cs Tracer Technique to the Estimate of Soil Erosion. Advance in Earth
Sciences, 15(5): 576–582.
Visser SM, Sterk G, Karssenberg D, 2005. Wind-erosion modelling in a
Sahelian environment. Environmental Modelling and Software,
20(1): 9–84
Walling DE, He Q, 1999. Improved models for estimating soil-erosion rates
from ceasium-137 measurements. J. Environmental Quality, 28(2):
611–622.
Walling DE, He Q, Whelan PA, 2003. Using 137Cs measurements to validate
the application of the AGNPS and ANSWERS erosion and sediment
yield models in two small Devon catchments. Soil & Tillage Research,
69: 27–43.
Woodruff NP, Siddoway FH, 1965. A wind-erosion equation. Soil Sci-
ence Society of America Proceedings, 29: 602–608.
Yan P, Dong GR, Zhang XB,Zhang YY, 2000. The preliminary result of
using 137Cs techniques to determine soil erosion of Qinghai-Tibet Plateau.
Chinese Sciences Bulletin, 45(2): 199–204.
Zhang CL, Zhou XY, Dong GR, Zhang XB, Qin Z, 2002. The 137Cs deposi-
tion features of arid-steppe area. Chinese Sciences Bulletin, 47(3):
221–225.
Zhang XB, Higgitt DL, Walling DE, 1990. A preliminary assessment of
potential for using cesium-137 to estimate rates of soil erosion in the
Loess Plateau of China. Hydrological Science Journal, 35: 243–252.
