Content uploaded by Paul F. Reich
Author content
All content in this area was uploaded by Paul F. Reich on Jun 09, 2014
Content may be subject to copyright.
This paper was peer-reviewed for scientific content.
Pages 838-846. In: D.E. Stott, R.H. Mohtar and G.C. Steinhardt (eds). 2001. Sustaining the Global Farm. Selected papers from the 10th International Soil
Conservation Organization Meeting held May 24-29, 1999 at Purdue University and the USDA-ARS National Soil Erosion Research Laboratory.
Global Dimensions of Vulnerability to Wind and Water Erosion
Paul Reich, Hari Eswaran and Fred Beinroth*
*Paul Reich and Hari Eswaran, USDA Natural Resources Conservation Service, Washington DC; Fred Beinroth, University of Puerto
Rico, Mayaguez, PR.*Corresponding author: fbeinroth@rumac.uprm.edu.
ABSTRACT
The global assessment of wind and water erosion
reported in this empirical study employs a simplified
model considering only soil and climatic variables. The
purpose is to obtain global and regional estimates of
areas of land susceptible to different intensities of
erosion. By relating to population densities, the areas of
high risk for water erosion are demarcated on maps and
they help to identify regions that require additional
attention to conservation. The actual intensity of erosion
is a function of many factors, the most important of
which is the level of management as determined partly
by the socioeconomic status of the land users. The
relationship between erosion vulnerability and the
inherent land quality (ILQ) is employed to compute
magnitudes of global soil loss through water erosion.
The annual potential yield of sediment through water
erosion from 72.5 million km2 of global land area
considered in this study is about 130 billion metric tons.
In the arable lands of the world, which are those
belonging to land quality classes I through VI, water
erosion may contribute about 67 billion metric tons of
sediment. In the susceptible drylands, which are the
areas most prone to desertification, water erosion could
yield about 92 Metric tons, which are about 71% of the
total global soil loss.
INTRODUCTION
Erosion is a natural land forming process that is
accelerated through human activities. Land surface
disturbance ranging from traffic moving over the land to the
extreme situation of deep tillage results in disruption of the
natural soil structural elements, loosening of particles, and
their eventual displacement. In the absence of chaotic events
such as sheets of water moving across a desert, water
induced erosion is minimal under natural conditions.
Erosion rates are strongly related to density and duration of
vegetative cover of the soil (Francis and Thorne, 1990).
With disturbance, onset of erosion is mostly a function of
slope and kinds of soil; the antecedent moisture content and
raindrop intensity are additional modifiers.
The art and science of soil conservation has developed
essentially around the central theme of managing wind and
water erosion. Numerous field experiments have been
reported in the literature and continue to be reported on
studies of evaluating water erosion under conditions of
different soil cover, slopes, and other variables. Assessments
of permanent micro-watersheds have yielded additional
information on the variables that affect the quantity and rates
of soil loss. Few attempts have been made to estimate soil
loss on watersheds or other large areas. The need to make
such estimates led to the development of the Universal Soil
Loss Equation (Wishmeier and Smith, 1978) and later the
Revised Universal Soil Loss Equation (Renard, 1997).
The first global assessment of human-induced water and
wind erosion was made by Oldeman et al. (1990) in the
framework of the Global Assessment of Land Degradation
(GLASOD) Project of the United Nations Environment
Program. Maps and information from this study, with minor
modifications are reproduced in the Atlas of Desertification
(Middleton and Thomas, 1997). Their estimates of land area
susceptible to wind and water erosion are given in Table 1.
In the GLASOD estimate, only the susceptible lands
from the dry sub-humid, semiarid, and arid parts of the
world were considered. Hulme and Marshe (1990) define
these climatic zones based on the Aridity Index (AI). AI is
the monthly sum of the ratios of precipitation to potential
evapotranspiration (P/PET) – dry sub-humid areas have a
ratio of <0.65, semiarid areas have a ratio of <0.5, arid areas
have a ratio of <0.2, and hyper-arid areas have a
Table 1. Distribution of susceptible dryland water and wind erosion in the continents. Source: Middleton and Thomas (1997).
WATER EROSION (million ha) WIND EROSION (million ha)
REGION Dry
Subhumid Semiarid Arid Dry
Subhumid
Semiarid Arid
Africa 25.1 59.2 34.8 1.6 30.7 127.5
Asia 54.9 69.9 32.7 15.1 52.1 85.9
Australasia 4.1 26.3 39.3 0 6.4 9.5
Europe 34.7 12.8 0.6 17.4 17.3 4.0
N. America 10.7 24.4 3.3 6.8 27.3 3.7
S. America 11.5 20.6 2.5 5.9 16.4 4.6
Total 141.0 213.2 113.2 46.8 150.2 235.2
TOTAL 467.4 432.2
P/PET ratio of <0.05. The susceptible drylands by
definition exclude the humid and hyper-arid areas as defined
by AI. The Aridity Index is a very crude assessment of
climate and errors are introduced by using this simplified
approach. Assessment of “human-induced erosion” was
made by opinions of a large number of persons, many of
whom had little or no actual data. Despite this, the
GLASOD assessment is the only estimate of human-induced
degradation.
PURPOSE
The purpose of this study is to obtain an estimate of
global land area, specifically in the susceptible drylands
(UNEP, 1992), vulnerable to water and wind erosion. El-
Swaify et al. (1982) called for “quantitative assessment of
erosion extent, tolerance limits, and causative parameters.”
Today there is a better understanding of factors controlling
erosion and estimates of acceptable levels under different
land uses. However, there is still an absence of erosion
estimates of national, regional or global extent. In addition
there is also an urgent need to locate critical erosion areas so
that strategies for priority action can be launched. The actual
rates of erosion are a function of on-site land quality
characteristics, climatic attributes, and the kinds of land
management technology that are applied. The management-
input level cannot as yet be quantified for a global level
assessment and is not considered in this analysis. However,
population density is used as a proxy to estimate this,
assuming that very high population densities are associated
with proportionately higher amounts of soil loss. Exceptions
to this occur in many of the developed countries of the
world, which have taken appropriate action to conserve the
soil resource.
METHODS
The current assessment considers ‘vulnerability to wind
and water erosion’ based on global soils and climatic data.
Using a geographic information system (GIS) a global soil
map of the world is used as the starting point of the
assessment. This map includes information on soil moisture
and temperature regimes. Rainfall induced erosivity is an
important parameter employed by several workers
(Charreau, 1969, Fournier, 1960, Delwaulle, 1973,
Wishmeier and Smith, 1978, and El-Swaify and Cooley,
1981) to estimate soil loss. Available data on erosivity was
related to soil moisture regimes to employ soil moisture
regimes as proxies. Based on the knowledge of soil behavior
under the prevailing climatic conditions, each soil unit was
assigned a vulnerability class to water and wind erosion.
Catastrophic events are excluded in this assessment. GIS
analysis was used to make global estimates of the area
occupied by each class.
Rates of soil loss under natural vegetation are generally
low and there is a steady state with very little variation with
time. Upon cultivation, the rates change dramatically.
Conservation technology such as implementation of
mechanical structures and/or zero-till or minimum tillage
reduces the amount of soil loss. However, a major factor is
the period when the land is bare or devoid of vegetative
cover and the number and intensity of storms at the onset of
the rainy season. Databases that provide such information
are just being assembled. In their absence, population
density in combination with soil and climatic attributes is
used to make estimates of rates of soil loss. This is done by
overlaying a population density map on maps depicting
vulnerability to wind and water erosion. A matrix (Table 2)
was developed to estimate the water erosion rate based on
population density and water erosion vulnerability.
Published data from countries of the world are used to arrive
at the assigned numbers. When no data were available,
values were assigned to the cells based on soil and climatic
attributes.
Table 2. Matrix to relate rates of erosion to population density
classes.
Annual Water Erosion Rates (Metric
tons/ha)
Vulnerability to Water Erosion
Population
Density
Persons/Sq.
km Low Mode
rate
High Very
High
<2 1 3 8 12
2-10 3 5 10 20
11-40 5 10 20 30
41-100 15 20 25 40
101-500 20 35 40 50
>500 25 50 75 80
Table 3. Matrix to relate rates of erosion to inherent land
quality classes.
Annual Water Erosion Rates (Metric tons/ha)
Vulnerability to Water Erosion
Inherent Land
Quality classes Low Moderat
e
High Very
High
I 2 4 6 8
II 3 5 8 11
III 4 8 12 17
IV 6 10 15 30
V 10 15 20 40
VI 15 20 40 60
VII 20 35 60 80
VIII 5 10 25 35
IX 1 4 6 10
Table 4. Comparison of land area in India impacted by
erosion in present study with that of Singh et al. (1992).
Area (‘000 km2)
Erosion Classes This study Singh et al.
(1992)
Low 870 801
Moderate 359 1,406
High 635 805
Very High 590 243
2,453 3,255
technology is applied. Even in the former systems, if the
land is used for padi cultivation during the rainy season or
The assumption in Table 2 is that if areas are highly
vulnerable to water erosion, high erosion rates result if the
population density is also high. This is true in low-input
low- output systems, but less valid where conservation
irrigation is available for cropping during the dry season, the
assumption is not correct. To assess soil loss as a function of
the inherent land quality classes (Eswaran et al., 1999), the
matrix of Table 3 was developed based on published erosion
rates in some countries. Due to the paucity of data, values
for most of the classes were interpolated to represent relative
magnitudes. This matrix is used later to compute magnitudes
of soil loss from water erosion.
To test the validity of the empirical model employed to
evaluate water erosion, data sources from India were used.
Singh et al., (1992) reported on erosion rates in India using
information from small runoff plots, small and large
watersheds, and sedimentation in reservoirs. Erosion rates
were measured with the Universal Soil Loss Equation
(Wishmeier and Smith, 1978) and the results were plotted on
a map. Table 4 compares the data of Singh et al. (1992) with
the estimates from this study. The study of Singh et al.
(1992) includes the very dry and cold areas of India, which
are excluded in the present study.
Results and Discussion
Table 5 is a summary of the susceptible dryland areas
subject to wind and water erosion on a global basis.
Globally there are about 19 million km2 (41% of susceptible
dryland) of land vulnerable to water erosion. The hyper-arid
and cold regions are excluded in this estimate. There are
about 17 million km2 or 37% of the drylands that are
vulnerable to wind erosion. The hyper-arid, cold, and humid
areas were excluded for the wind erosion estimate.
According to the GLASOD (Oldeman et al., 1990) estimate,
8.3% of the susceptible dryland areas are degraded due to
human-induced water erosion and 13.2% experience
accelerated wind erosion due to human interventions.
Table 6 is a summary of the global land areas subject to
wind and water erosion. Globally there are about 56 million
km2 (43% of ice-free land) of land vulnerable to water
erosion in the slightly arid to humid areas of the world. The
hyper-arid and cold regions are excluded in this estimate.
There is about 33 million km2 or 25% of the landmass that is
vulnerable to wind erosion. The hyper-arid, cold, and humid
areas were excluded for the wind erosion estimate.
The distribution of the lands vulnerable to water and
wind erosion is shown in Figs. 1 and 2. Lands of high to
very high vulnerability, those most susceptible to erosion,
Table 5. Estimates of susceptible dryland areas vulnerable to water and wind erosion. See
Middleton and Thomas (1997) for susceptible dryland definition.
WATER EROSION WIND EROSION
EROSION VULNERABILITY
CLASS AREA
Million km2 Percent AREA
Million km2 Percent
LOW 5.15 11.2 3.64 7.9
MODERATE 5.76 12.5 3.74 8.1
HIGH 3.31 7.2 5.92 12.9
VERY HIGH 4.44 9.6 3.75 8.1
Total Vulnerable Area 18.66 40.5 17.05 37.0
Dry 23.38 50.8 23.17 50.3
Depositional/minimal 3.92 8.5 5.75 12.5
Cold 0.06 .1 0.06 .1
Total Susceptible Drylands 46.02 100 46.02 100
Table 6. Estimates of global land areas vulnerable to water and wind erosion. Note: the humid
areas were included for the water-erosion estimate but not for the wind erosion. The 40 million
km2 of the humid areas are assumed to have negligible wind erosion
WATER EROSION WIND EROSION
EROSION
VULNERABILITY CLASS AREA
Million km2 Percent AREA
Million km2 Percent
LOW 17.33 13.3 9.25 7.1
MODERATE 15.39 11.8 6.32 4.8
HIGH 10.97 8.4 7.80 5.9
VERY HIGH 12.21 9.3 9.33 7.1
Total Vulnerable Area 55.91 42.8 32.70 25.0
Dry 37.77 28.9 37.77 28.9
Depositional/minimal 16.59 12.7 40.00 30.6
Cold 20.36 15.6 20.36 15.6
Total Ice-free Land Area 130.63 130.63
Figure 1. Water erosion vulnerability.
Figure 2. Wind erosion vulnerability.
Figure 3. Risk of human induced water erosion.
Table 7. Vulnerability to erosion in relation to inherent land quality. Note: The extreme drylands in ILQ class IX and
extreme cold lands in ILQ class VIII are excluded.
WATER EROSION (Million km2) WIND EROSION (Million km2) ILQ
CLASS LOW MODERATE HIGH V. HIGH LOW MODERATE HIGH V. HIGH
I 1.22 1.41 1.39 0.02 0.72 1.04 1.00 0.02
II 1.43 1.19 1.85 1.82 0.08 0.60 0.25 0.34
III 2.42 1.69 0.41 0.17 1.29 0.80 0.70 0.07
IV 0.07 0.47 0.03 0 0.03 0.04 0.01 0
V 7.45 4.46 6.15 2.21 6.56 3.12 2.92 1.05
VI 4.57 6.01 0.89 0 0.47 0.58 2.82 0
VII 0 0.01 0 7.35 0 0 0 7.35
VIII 0.05 0.14 0.26 0.07 0.05 0.14 0.02 0.04
IX 0.12 0.01 0.01 0.55 0.04 0.01 0.06 0.05
Total 17.33 15.39 10.97 12.21 9.25 6.32 7.80 9.33
Table 8. Estimate of global annual water erosion soil loss as a function of land quality classes of potentially arable
lands. LQI VII, VIII, and IX are not included in this table.
Annual Water Erosion Amount (billion Metric tons)
Vulnerability to Water Erosion
Land Quality Class
Low Moderate High Very High TOTAL
I 0.24 0.56 0.83 0.02 1.66
II 0.43 0.6 1.48 2.01 4.51
III 0.97 0.14 0.21 0 2.53
IV 0.04 0.47 0.07 0 0.58
V 7.45 6.69 12.29 8.85 35.28
VI 6.86 12.02 3.52 0 22.40
TOTAL 15.99 21.70 18.40 10.88 66.96
occur mostly in the poorer third world countries. In Africa,
they are dominant in the region between the southern limit
of the Sahara and the humid zone. In South Asia, the whole
of India is highly vulnerable. India, in general, is comprised
of an old geomorphic surface, uplifted and tilted with a
current highly erosive climate. These physiographic
conditions, coupled with a farming system that is essentially
low-input with few conservation technologies, leads to high
rates of human-induced water erosion. Erosion, with rates
exceeding 30 Metric tons/ha/yr (Singh et al., 1999) is so
rampant in the central part of the continent that large
inselbergs are exposed in a north-south alignment. The
highest erosion rates are probably along the east-west
running Shiwalik hill tracts (estimated by Singh et al. (1992)
as being about 80 Metric tons/ha/yr) but here it is largely
coupled with tectonic uplifts.
Though peneplanation processes formed the large
expanses of the Mid- to End-Tertiary erosion surfaces in
Central and Southern Africa, they are not currently sites of
intensive erosion. The flat nature of the present day surfaces
does not promote intensive erosion under the prevailing
climate. During the period of peneplanation, both the
physiographic conditions and the climate must have been
different and were perhaps similar to those in India today. It
is difficult to validate a global assessment. However, it is
possible to compare the results obtained in this study with
estimates obtained for specific locations published in the
literature. A synthesis of such data was compiled by El-
Swaify et al. (1982) and is a valuable source of information
for comparison. Due to insufficient national estimates, the
validation was only made for water erosion susceptibility.
To estimate the amounts of soil loss due to water
erosion, land quality classes (Eswaran et al., 1999) were
assigned erosion rates for each of the vulnerability classes.
As indicated in the methodology, the amounts assigned
(Table 3) were empirical values obtained from studies in
different parts of the world. It is estimated by this procedure
that the total global soil loss, in all lands excluding the
extreme dry and the extreme cold, which is about 72.5
million km2, through water erosion is about 130 Metric tons
per year. Table 8 presents the estimates for each land quality
classes (only I through VI), which represents most of the
arable lands of the world. The total estimate of soil loss due
to water erosion for the six classes is about 67 billion Metric
tons. This is about 52% of the global annual soil loss.
As seen in Table 8, very high erosion losses arise from
Class V and VI lands, which together account for over 86%
of total soil loss from the arable lands. Much of Class V and
VI lands occur in the inter-tropical areas that have high
population densities and low inputs for land management as
they can least afford appropriate conservation technology.
Class V and VI lands are in general fragile ecosystems and
also form the main challenges to institute sustainable
agriculture. These lands are also one of the major sources of
carbon emission to the atmosphere, which of course is
enhanced through erosion.
In the United States, the US Department of Agriculture
Natural Resources Conservation Service (USDA, 1996)
Table 9. Global soil loss due to water erosion in relation to population density. Note that very cold and dry areas
are excluded from the estimate. About 16.6 million km2 of land (Table 2) is considered as depositional and
contributes minimally to water erosion.
Erosion amount (million Metric tons)
Vulnerability to erosion
Population Density
Persons/Sq. km Low Moderate High Very High Total
<2 392 842 717 1,935 3,886
2-10 1,633 2,247 2,429 15,341 21,650
11-40 1,878 4,092 5,589 9,978 21,537
41-100 3,025 4,465 5,116 8,696 21,302
101-500 3,460 5,280 9,636 10,375 28,751
>500 1,029 873 2,148 1,779 5,829
Total 11,418 17,799 25,634 38,103 91,953
estimated that in 1982, 3.14 billion metric tons of soil were
lost through wind erosion and 3.8 billion metric tons by
water erosion. This total soil loss of about 7 billion metric
tons was reduced through the Conservation Reserve
Program (CRP) to about 5 billion metric tons by 1992.
These estimates are only for cropland. In our study of the
US, we estimated the potential soil loss through water
erosion for land belonging to land quality Classes I through
VI in the rating of Eswaran et al. (1999). The potential soil
loss is about 6.5 billion metric tons. The actual soil loss of
about 3.1 billion metric tons is about 48% of our estimated
potential soil loss and this points to the success of
implementing a conservation technology program. In India,
the amount of soil loss computed from the data of Singh et
al. (1992) is about 4.7 billion metric tons for the whole of
India. In our study, we estimate about 4.3 billion metric tons
of soil loss by water erosion for the area of India under
consideration.
To demarcate risk of water erosion arising from land use,
we overlaid the water erosion vulnerability map with a
population density map. An estimate of amounts of soil loss
that can be expected for each population density class is
obtained. Fig. 3 shows the areas of high risk. Using the
matrix in Table 2, the amounts of potential soil loss were
calculated and presented in Table 9.
Table 9 suggests that the annual soil loss due to water
erosion, in the regions of the world prone to desertification
(55.91 million km2), is about 92 billion Metric tons. The
average rate is about 16 metric tons ha-1 yr-1. Fig. 3 shows
the distribution of these lands. Again, the major
concentration of land with high erosion rates or erosion risks
is in the poorer third world countries.
CONCLUSION
Lowdermilk (1953) has provided a good historical
analysis of the rise and fall of civilizations and how this was
related to the quality of land resources. Though there is good
evidence for enhanced soil loss in poorer countries of the
world there is no good estimate for global soil loss through
wind and water erosion. The GLASOD (Oldeman et al.,
1990) project recently, made the first attempt to evaluate
human-induced land degradation. The land areas with high
erosion rates and their locations are already disheartening as
they confirmed previous published observations.
In this analysis, we preferred to consider wind and water
erosion more generically and attempted to locate vulnerable
areas and provide estimates of global magnitudes. The
assessment is necessarily empirical, since no global database
exists. Therefore, assessments are largely dependent upon
assumptions made in the method. The results, however,
validate the widely held belief of the extensive nature of the
problem and its seriousness. The 56 million km2 of land
vulnerable to water erosion has the capacity to yield about
92 billion metric tons annually of sediment that can have
significant off-site damages. The on-site damage ranges
from reduced soil quality to collapse of biodiversity and
ecosystems. In many of the third world countries, food
security is strongly linked to quality of land resources and
these are the countries in risk of famine and civilian unrest
due to extensive land degradation.
It is acknowledged that the weakness of this analysis
stems from the lack of reliable data and consequently the use
of assumptions that cannot be unequivocally proven. In the
absence of national assessments and monitoring programs,
we may have to make such judgments based on default
values. This study offers evidence that soil erosion and the
resulting degradation remains a threat to world food
production. Accordingly, a global program to make
assessments and monitor this process is urgently needed.
REFERENCES
Charreau, C. 1969. Influence des techniques culturales sur le
developpement du ruissellement et de l’erosion en
Casamance. Agron. Trop. 24:836-844.
Delwaulle, J.C. 1973. Resultats de six ans d’observations sur
l’erosion ai Niger. Revue Bois et Forets. 150:15-37.
El-Swaify, S.A. and K.R. Cooley. 1981. Soil losses from
sugarcane and pineapple land in Hawaii. In: De Boodt,
M and A. Gabriels (eds.) 32-340.
El-Swaify, S.A., E.W. Dangler and C.L. Armstrong. 1982.
Soil Erosion by Water in the Tropics. Univ. of Hawaii,
Research Extension Series 024. 173 pp.
Francis, C.F. and J.B. Thornes. 1990. Runoff hydrographs
from three Mediterranean vegetation cover types. In:
Thornes, J.B. (ed.) Vegetation and Geomorphology.
Chichester Wiley, England. 363-385.
Hulme, M. and R. Marsh. 1990. Global mean monthly
humidity surfaces for 1930-59, 1960-89, and projected
for 2020. Report to UNEP/GEMS/GRID. Climatic
Research Unit, University of East Anglia, Norwich,
England.
Lowdermilk, W.C. 1953. Conquest of the land through
seven thousand years. Agric. Info. Bull. No. 99. USDA.
Washington, DC.
Middleton, N and D, Thomas. 1997. World Atlas of
Desertification. Published for UNEP by Arnold Publ.
2nd. Edition. London. 182 pp.
Oldeman, L.R., R.T.A. Hakkeling and W.G. Sombroek.
1990. World map of the status of human-induced soil
degradation: An Explanatory note. ISRIC. Wageningen,
Netherlands.
Renard, K.G. 1997. Predicting soil erosion by water : a
guide to conservation planning with the revised universal
soil loss equation (RUSLE). Agriculture Handbook; No.
703. Washington, D.C., USDA, Agricultural Research
Service, 384 p.
Singh, G., R. Babu, P. Narain, L.S. Bushan and I.P. Abrol.
1992. Soil erosion rates in India. J. Soil and Water Cons.
47:97-99.
USDA. 1996. A Geography of Hope. Publ. US Department
of Agriculture, Natural Resources Conservation Service.
80 pp. Washington DC.
Wishmeier, W.H. and D.H. Smith. 1978. Predicting rainfall
erosion losses – a guide to conservation planning. Agric.
Handbook No. 537. USDA, Washington, DC.
