Printer Friendly

Modeling wind and water erosion in Northern China under climate and land use changes.

**********

The semi-arid and arid region of northern China has suffered from high soil loss rates due to wind and water over the past century, resulting in gradually enlarged deserts (Brown 1984; Higgitt 1993; Zhang and Shao 1997; Zhang, Quine, and Walling 1998). This problem has been exacerbated by increased economic pressure due to increased population over the last few decades. Research has been conducted to identify causes of desertification for this region. While many factors contributed to desertification, changes in land use from forests, shrublands, and grasslands into croplands (Dregne 1992) and climate (Williams et al. 1996) are believed to be two major factors affecting soil erosion and degradation of land and vegetation.

Reversing effects of desertification requires (1) quantitatively understanding mechanisms governing soil erosion, (2) identifying root causes of land degradation, (3) predicting the amount and distribution of soil loss in relation to possible causal factors, and (4) making erosion assessment for conservation policy-making (Heimlich and Bills 1984; Perrens and Trustrum 1985). Erosion by wind and the protective role of sparse vegetation against wind erosion have been studied in great detail (Wolfe and Nickling 1993; Wasson and Nanninga 1986). Based on results of extensive wind tunnel experiments on the effects of vegetation coverage on sand transportation, Buckley (1987) proposed a model for wind erosion

q=B[[V(1-kC)-[V.sub.t]].sup.3] (1)

where

q is the sand transport (g [cm.sup.-1] [s.sup.1]).

B is a constant (g [s.sup.-2] [cm.sup.-1] [m.sup.-2]) related to soil cohesion.

V is the wind velocity (m [s.sup.-1]) at 0.5 m elevation.

[V.sub.t] is a threshold wind velocity (m [s.sup.-1]).

C is vegetation coverage up to 17%.

k is a dimensionless constant depending on plant geometry

Revised wind erosion equation (Fryrear et al. 1998; Merrill et al. 1999) is a typical empirical model developed to predict wind erosion, whereas Wind Erosion Prediction System (WEPS) represents more mechanistic process-based approaches (Hagen 1991).

Water erosion models can be classified into two categories, empirical models and process- based mechanistic models. One empirical model is the Universal Soil Loss Equation (USLE) developed and validated mainly in the United States (Wischmeier and Smith 1965; Wischmeier 1976; Bengtson and Sabbagh 1990), and later modified (Kinnell and Risse 1998) and revised (Renard et al. 1991, 1997; Renard and Freimund 1994; Yoder and Lown 1995).These modeli have a simple mathematical structure, with soil loss rate equal to a product of a number of factors such as soil erodibility, runoff erosivity, slope steepness and length, vegetation coverage, and conservation practice. Field erosion data are used to obtain parameters in the models. While these models have the advantage of being mathematically simple and have been used extensively for practical purposes of conservation planning, their empirical nature makes it difficult to predict accurately some local and site-specific conditions. Mechanistic process-based models, such as Water Erosion Prediction Project (WEPP) and Limburg Soil Erosion Model (LISEM) (Laflen et al. 1991; Panuska et al. 1991; Laguna and Giraldez 1993; Smith et al. 1995; Tiscareno-Lopez et al. 1995; Roo, Offermans, and Cremers 1996; Roo, Wesseling, and Ritsema 1996; Parsons et al. 1997; Botterweg et al. 1998), represent a new generation of soil erosion models. Water erosion was separated into nil and interrill erosion (Abrahams et al. 1989; Loch 1996), and the process of soil erosion was further divided into soil detachment, sediment transport and deposition. Interrill soil loss increased with rainfall intensity or raindrop energy but decreased with surface flow depth. In contrast, nil soil loss increased with flow velocity and flow depth and depended on nil geometry (Fentie et al. 1997). Sediment transport and deposition were modeled with mass balance equations and sediment carrying capacity, which depends on flow velocity. Unlike empirical models, factors such as slope, vegetation, and soil erodibility were implicitly included in the flow amount, flow velocity, and threshold soil shear stress in these mechanistic models. Compared to empirical models such as USLE and Revised Universal Soil Loss Equation (RUSLE), process-based models are more theoretically sound. However, with more complicated mathematical structure and detailed consideration of local geometry and physical processes, application of these models is generally limited to relatively small spatial-temporal domains. Some soil erosion mechanisms, such as the role of vegetation on soil erosion (Weltz et al. 1998), are still not well understood; therefore, results may be limited. Our objective was to simulate soil erosion by wind and water in a sandy grassland in Inner Mongolia, China, using a semi-empirical approach. The effect of vegetation, represented by Advanced Very High Resolution Radiometry (AVHRR) vegetation index by satellite remote sensing, on wind and water erosion was included in our model. Simulations were run for combinations of land use cha nges and climate scenarios to provide predictions for potential soil erosion response in the sandy grassland.

Methods and Materials

Site description. Our study site was Yijinhuoluo County in Inner Mongolia Autonomous Region of China, located in the east of Erdos Plateau, and ranging from 109.0037[degrees]-110.5037[degrees] E and from 39.9054[degrees]-39.8054[degrees] N, with a total area of 5891 [km.sup.2] (2275 [mi.sup.2]). Annual mean temperature is about 7[degrees]C (44.6[degrees]F), with 20 and -10[degrees]C (68 and 14[degrees]F) as the warmest (July) and coldest (January) monthly means. Annual precipitation is about 350 mm (13.8 in),, with more than 60% occurring in July and August. Winds are generally from the northwest in winter and spring, when daily mean wind speed can be more than 10 in [s.sup.-1] (32.8 ft [s.sup.-1]). Major climate gradients run from southeast to northwest, with gradually decreasing temperature and precipitation but increasing wind speed from the southeast toward northwest. The county has hilly landscapes and sandy soils. The dry and windy climate and hilly sandy landscapes result in wind erosion in most areas of the county. Calcium-rich soils in the eastern part of the county are more difficult to erode by wind, but are more susceptible to water erosion. Soil loss via rills and gullies is the dominant form of erosion in the eastern part of the county (Zhang and Shao 1997; Zhang, Quine, and Walling 1998). Figure 1 shows basic information for the county.

Model description. The following assumptions were made during mathematical formulation:

(1) Wind erosion mainly occurs on hilly sandy soils in uplands, and wind detaches topsoil particles and transports them to lower areas with rills and gullies.

(2) Water erosion occurs mainly in lower areas as rills and gullies, and runoff is a major contributor of soil detachment.

(3) Compared to runoff, soil detachment by raindrops is negligible. At the field scale, it is difficult to separate out effects due to raindrop impact versus effects due to runoff, or "stream power." At larger scales, runoff appears to be more dominant than raindrop impact (Zhang, Quine, and Walling 1998).

(4) Runoff in rills and gullies can effectively remove all detached soil so sediment transport does not need to be explicitly described mathematically. In other words, simulation of the steady-state soil loss from the system was attempted, not the dynamic process of detachment, transportation, and deposition.

Our model for wind erosion was similar to that of Buckley (1987), with a modification to include the effects of soil water content and vegetation index on sand transport

[E.sub.wind] = max{0,[C.sup.J.sub.wind][{ll-[[ll.sub.0] + [U.sup.J.sub.w] w/[W.sup.j.sub.0] + [U.sup.J.sub.v]max[0,([V.sup.i.sub.j]-[V.sub.0])]]}.sup.3]} (2)

where

[E.sub.wind] is the daily total soil loss by wind (t [km.sup.-2] [d.sup.-1], or 2.548 ton [mi.sup.-2] [d.sup.-1]).

[C.sub.j.sub.wind] is a coefficient related to inherent erodibility of soil type j (t [km.sup.-2] [d.sup.-1] [s.sup.3] [m.sup.-3], or 0.0722 ton [mi.sup.-2] [d.sup.-1] [s.sup.3] [ft.sup.-3]).

u is the daily mean wind speed by meteorological stations (m [s.sup.-1], or 3.28 ft [s.sup.-1]).

[u.sub.0] is threshold wind velocity for bare soil with zero water content.

w is the daily mean volumetric soil water content in top soil.

[W.sup.j.sub.0] is the volumetric soil water content at saturation for soil type j.

[U.sup.j.sub.w] is a coefficient signifying the reduction effect of soil water on soil loss for soil type j (m [s.sup.-1], or 3.28 ft [s.sup.-1]).

[V.sup.i.sub.j][member of] [0,1], is AVHRR vegetation index on soil type j with land use type i.

[V.sub.0] is an offset value of [V.sup.i.sub.j] for zero vegetation, and was estimated by averaging the vegetation index over the non-vegetated surfaces.

[U.sup.j.sub.v] is a coefficient signifying the reduction of soil erosion by vegetation (m [s.sup.-1], 3.28 ft [s.sup.-1]).

We used remote sensing AVHRR vegetation index here because there were significantly positive relationships between the vegetation index and green biomass (Tucker 1979). Equation 2 implies that the daily wind erosion is proportional to the difference between daily mean wind speed and a critical wind speed raised to the third power. Daily soil erosion decreases with vegetation coverage and soil water content because of increased critical wind speed.

Combining empirical knowledge presented in USLE and RUSLE (Renard et al. 1991; Renard and Freimund 1994; Yoder and Lown 1995), and mechanisms of water erosion described in WEPP and LISEM (Laflen et al. 1991; Roo, Offermans, and Cremers 1996; Roo, Wesseling, and Ritsema 1996), we formulated water erosion rate as a function of surface flow, slope steepness, and vegetation status. The function takes the following form

[E.sub.water] = [C.sup.j.sub.water] [(R/[R.sub.0]).sup.[beta]] (1+[(S/[S.sub.0]).sup.[sigma]])exp[-[alpha]max(0,[V.sup.i.sub.j]-[V.s ub.0])] (3)

where

[E.sub.water] is the daily erosion by runoff (t [km.sup.-2] [d.sup.-1], 2.548 ton [mi.sup.-2] [d.sup.-1]).

[C.sup.j.sub.water] is a coefficient (t [km.sup.-2] [d.sup.-1], or 2.548 ton [mi.sup.-2] [d.sup.-1]).

R is the amount of surface flow (mm, 0.0394 in).

S is the steepness of slope (m [m.sup.-1], ft [ft.sup.-1]).

[R.sub.0] and [S.sub.0] are constants to make the terms of surface flow and slope dimensionless.

exponent [beta] was empirically determined as 0.5, and [sigma] was similarly chosen as 1.5.

[alpha] is a shape parameter describing the protective role of vegetation for soil against rill erosion.

Equation (3) assumes that rill erosion increases with overland flow and slope steepness, but decreases with vegetation index. Effect of slope length is partially included in R because longer slopes result in more accumulated surface flow.

The daily increment of soil water was computed as precipitation plus runon from up-gradient areas minus evapotranspiration and runoff

[DELTA]w = 1/[D.sub.j] (p+runon-runoff - [E.sub.p] w/[W.sup.j.sub.0]) (4)

where

[DELTA]W is the daily increment of volumetric soil water content.

[D.sub.j] is the thickness of the effective top soil layer (mm).

p is the daily rainfall amount (mm).

runon is the inflow from adjacent area, or 4 neighborhood gird cells, with higher elevation (mm).

runoff is the locally produced runoff (mm). [E.sub.p] is the daily potential evaportranspiration (mm) computed from Penman's equation (Rosenberg et al. 1983).

Runoff is produced at a local spot when top soil is saturated, i.e.,

[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]

Runoff or outflow from a local point (grid cell) is distributed to lower neighborhood grid cells so that the lowest neighboring grid cell receives the largest amount of runoff. In particular, if a grid cell at row m and column n is denoted by [g.sub.(m,n)] and its elevation by [h.sub.(m,n)], and flow out of grid cell [g.sub.(m,n)] is distributed its four neighbors [g.sub.(m+1,n)], [g.sub.(m-1,n)], [g.sub.(m,n+1)], [g.sub.(m,n -1)] using the following equation

[runon.sub.(m,n)[right arrow](m+1,n)] = [e.sub.(m+1,n)] [[h.sub.(m,n)] - [h.sub.(m+1,n)]] [runoff.sub.(m,n)]/[summation over (p=-1,1)] [e.sub.(m+p,n)] [[h.sub.(m,n)] - [h.sub.(m+p,n)]] + [summation over (q=-1.1)] [e.sub.(m,n+q)] [[h.sub.(m,n)] - [h.sub.(m,n+q)]]

[runon.sub.(m,n)[right arrow](m-1,n)] = [e.sub.(m-1,n)] [[h.sub.(m,n)] - [h.sub.(m-1,n)]] [runoff.sub.(m,n)]/[summation over (p=-1,1)] [e.sub.(m+p,n)] [[h.sub.(m,n)] - [h.sub.(m+p,n)]] + [summation over (q=-1.1)] [e.sub.(m,n+q)] [[h.sub.(m,n)] - [h.sub.(m,n+q)]]

[runon.sub.(m,n)[right arrow](m+1,n)] = [e.sub.(m,n+1)] [[h.sub.(m,n)] - [h.sub.(m,n+1)]] [runoff.sub.(m,n)]/[summation over (p=-1,1)] [e.sub.(m+p,n)] [[h.sub.(m,n)] - [h.sub.(m+p,n)]] + [summation over (q=-1.1)] [e.sub.(m,n+q)] [[h.sub.(m,n)] - [h.sub.(m,n+q)]] (6)

[runon.sub.(m,n)[right arrow](m,n-1)] = [e.sub.(m,n-1)] [[h.sub.(m,n)] - [h.sub.(m,n-1)]] [runoff.sub.(m,n)]/[summation over (p=-1,1)] [e.sub.(m+p,n)] [[h.sub.(m,n)] - [h.sub.(m+p,n)]] + [summation over (q=.1.1)] [e.sub.(m,n+q)] [[h.sub.(m,n)] - [h.sub.(m,n+q)]]

where

[rumon.sub.(m,n)[right arrow](i,j)] is the flow from [g.sub.(m,n)] to [g.sub.(i,j)] [e.sub.(ij)] is zero if [h.sub.(i,j)] is larger than [h.sub.(m,n)] or [g.sub.(i,j)] is out of the boundary of the spatial simulation domain, otherwise [e.sub.(i,j)] is 1.

[runoff.sub.(m,n)] is the runoff from [g.sub.(m,n)].

Here free index variables are I = m+1 or m-1, and j = n+1 or n-1. Equations 4-6 describe a recursive definition and computation of runoff and soil water.

Data sets and model validation. The following data sets were used in this study: (1) Elevation, land use, soil classification, and soil erosion maps of Yijinhuoluo County (Figure 1); (2) Monthly AVHRR vegetation index maps from l992-l996 from NOAA satellite; (3) Maps of monthly precipitation, temperature, wind speed, relative humidity (Figure 1) and cloudiness averaged over 1981-1990; and (4) 10-year daily climate records from January 1, 1989 to December 31, 1998. The land use and soil classification maps were compiled in the late 1980s. Soil erosion was mapped by Inner Mongolia Soil and Water Conservation Institute in 1998 by extrapolating field measurements/estimations in light of infrared air photography (Wang et al. 1992). Daily climate records included precipitation; maximum, minimum, and mean temperatures; mean relative humidity; mean wind speed; and mean cloudiness.

All maps were digitized into 0.00833333[degrees] latitude by 0.00833333[degrees] longitude resolution, making each grid cell about 926.6 m by 716.5 ix (3039 ft by 2350 ft). There were 8873 nonempty grid cells. Soils were grouped into six classes named calcium soil (CALC), hilly sandy soil (HSND), meadow sandy soil (MSND), gravel soil (GRAV), wetland soil (WETL), and non-erodible surfaces (OTHR). Similarly, land uses were also grouped into seven categories: sand dunes (DUNE), dry crop fields (DCRP), irrigated crop fields (WCRP), forests (TREE), shrublands (SHRB), grasslands (GRAS), and other non-erodible land use (NERO). Table 1 gives the number of grid cells in each soil and land use category. Surface thickness of each soil was derived from the thicknesses indicated on the soil map.

Excluding the OTHR soil class and NERO land use class, 30 joint classes were obtained (Table 1). Vegetation index in August (1992-1996) was averaged for each joint class to give [V.sup.i.sub.j] for [i.sup.th] land use type and [j.sup.th] soil class in Equations 2 and 3. Similarly, wind and water erosion (t [km.sup.-2] [yr.sup.-1], 2.548 ton [mi.sup.-2] [yr.sup.-1]) (Figure 1) were integrated for each joint class to give the total wind and water erosion (t [yr.sup.-1], 0.984 ton [yr.sup.-1]) of the class, These class totals were compared to simulated class totals to provide a partial validation. Since daily climate data were available for one point (the county meteorology station) only, we scaled these daily climatic quantities (precipitation; mean wind speed; maximum, minimum, and mean temperature; mean relative humidity; and mean cloudiness) in light of corresponding monthly distribution maps over the county to produce daily climatic maps by means of the following expression

[V.sub.CD](x,y) = [V.sub.CD]([x.sub.0],[y.sub.0]) [V.sub.CM](x,y)/[V.sub.CM]([x.sub.0],[y.sub.0]) (7)

where

[V.sub.CD](x,y) and [V.sub.CM](x, y) are respective daily and monthly climatic variables [V.sub.C] ([V.sub.C] can be precipitation, mean temperature, mean wind speed, mean relative humidity, or mean cloudiness) at point (x,y).

x and y denote longitude and latitude, respectively.

([x.sub.0],[y.sub.0]) is the location of the county meteorological station

Maps of [V.sub.CD](x,y) were then used as input data for the model.

Parameters in Equations 2 and 3, i.e., [C.sup.j.sub.wind], [C.sup.j.sub.water], [U.sup.i.sub.v] and [U.sup.i.sub.iv], were empirically chosen and adjusted based on soil and vegetation properties. Simulations were run for each grid cell using Equations 2-6 and daily climatic data (1989-1998). The time step of integration was one day. Average annual wind and water erosion over the 10 years was then obtained from simulated daily erosions for each grid cell. Simulated annual erosion was integrated spatially for each joint soil-land-use class and compared to corresponding values obtained from the soil erosion map.

Simulations for climate and land use changes. To investigate effects of climatic shift on potential soil loss, we examined temporal patterns of climatic variables from 1959-1998. Statistical analysis indicated that annual mean temperature has increased 0.956[degrees]C over the past 40 years, but no statistically significant change was observed for other variables (Gao 1999). Hence, we designed four climatic scenarios, T0, T1, T2, and T3, by introducing uniform temperature increases of 0, 1, 2, and 3[degrees]C, respectively, over the region and the 10-year period, while other climatic variables remained unchanged.

We were also interested in how soil erosion changes with incremental changes in land use patterns based on our model. Five land-use scenarios, L0, L1, L2, L3, and L4, were produced by

* keeping the present land use pattern

* converting grasslands of wetland soils (WETL) into dry crop fields

* converting grasslands of calcium and wetland soils (CALC + WETL) into dry crop fields

* converting grasslands of calcium soils, wetland soils, and meadow sandy soils (CALC + WETL + MSND) into dry crop fields

* converting all grasslands into dry crop fields

Twenty joint climate-land-use scenarios were produced by crossing four temperature scenarios and five land use scenarios. Scenarios were then used in the model over 10-year simulations to obtain individual and interaction effects of these two factors on soil loss.

Results and Discussion

Model validation. Comparisons between data from the erosion map and simulated total annual wind and water erosion for each of the 30 joint soil-land-use categories are given in Figure 2. In most cases, model results agreed with data, except for a few grid cells where overestimations of erosion occurred for irrigated crops (WCRP) and forest (TREE) in calcium soil (CALC) and forest in gravel soil (GRAV). Overestimations of water erosion occurred for dry crops (DCRP) in all soils except gravel soil (GRAV). Underestimated erosion occurred for irrigated crops (WCRP) in wetland soil (WETL). Overprediction of water erosion was partially due to "hidden" water erosion in areas where wind erosion was predominant. Water erosion for these areas was not recorded on the soil erosion map and was neglected in the integration for class totals.

With observed climate data and unchanged land use pattern, simulated wind erosion (EV00 in Figure 3) varied from about 1000 to more than 30.000 t [km.sup.-2] [yr.sub.-1] (25480 to 75440 ton [mi.sup.-2] [yr.sup.-1]), and the predicted wind erosion in the western part of the region was 30 times greater than in the eastern part of the county. Soil in the western part of the region was sandier and the climate drier and windier than the eastern part of the region. On the other hand, about 80% of the area where soil loss by water was greater than 750 t [km.sup.-2] [yr.sup.-1](1911 ton [mi.sup.-2] [yr.sup.-1]), was in the eastern part of the region (EW00 in Figure 4). Greater water erosion in the east was due to higher precipitation and steeper slopes compared to the western part of the county (Figure 1). As a result, more frequent rill flow with greater stream power was simulated in the eastern part (not shown). Spatial distribution of simulated soil loss agreed with spatial trends of the erosion map. Hence, our m odel was partially validated and justified for this case study.

Simulations for temperature and land use changes. Simulated wind and water erosion for four temperature scenarios with unchanged land use patterns and four land use scenarios with unchanged temperature, are shown in Figures 3 and 4. Figure 5 provides average annual soil loss for each of the 20 joint scenarios. Simulation results for present climate and unchanged land use (L0 and T0) gave 6552 and 1125 t [km.sup.-2] [yr.sup.-1] (2571 and 2866 ton [mi.sup.-2] [yr.sup.-1] as average wind and water erosion, respectively. Combined average erosion (wind + water) was thus 7677 t [km.sup.-2] [yr.sup.-1] (19561 ton [mi.sup.-2] [yr.sup.-1]). If average soil bulk density is about 1.5 t [m.sup.-3] (0.04251 ton [ft.sup.-3]), this combined average erosion rate means the county lost on average 5 mm of topsoil every year. Increasing temperature increased wind erosion. Higher temperatures increase evapo-transpiration and reduce soil water content, making soil more erodible by wind (D_EV10, E_EV20 D_EV30 in Figure 3). Larger i ncreases in wind erosion occurred in the western part of the county because soil there is sandier than in the eastern part. Conversely, water erosion decreased with increased temperature (D_EW10, D_EW20 D_EW30 in Figure 4). Decreases in soil water content with increased temperature caused less runoff during rainfall events, thus descreasing rainfall erosivity. Decrease in water erosion was more evident in the eastern part than in the western part of the county because of steeper slopes and higher precipitation in the east, With land use remaining unchanged, the average simulated wind erosion for the county increased from 6552 to 6647 t [km.sup.-2] [yr.sup.-1] (16694 to 16936 ton [mi.sup.-2] [yr.sup.-1]). However, average water erosion decreased from 1125 to 1112 t [km.sup.-2] [yr.sup.-1] (2867 to 2833 ton [mi.sup.-2] [yr.sup.-1]), when temperature was increased by 3[degrees]C (5.4[degrees]F). The overall effect of increased temperature was a net positive increase of 82 t [km.sup.-2] [yr.sup.-1] (209 ton [mi.s up.-2] [yr.sup.-1]) in average annual soil loss.

Wind and water erosion increased for all land use change scenarios. Increase in wind erosion was larger in the western part of the county than in the east (D_EV01 D_EV02, D_EV03 and D_EV04 in Figure 3), and areas with erosion greater than 750 t [km.sup.-2] [yr.sup.-1] (1911 ton [mi.sup.-2] [yr.sup.-1] mainly occurred in the western part. Increase in water erosion due to land use changes occurred mainly in the eastern part of the county (D EW01, D_EW02, D_EW03, and D_EW04 in Figure 4). With temperature unchanged, average wind erosion changed from 6552 t [km.sup.-2] [yr.sup.-1] (16694 ton [mi.sup.-2] [yr.sup.1]) for present land use patterns to 6681, 6786, 6817, and 7392 t [km.sup.-2] [yr.sup.-1] (17023, 17290, 17369, and 18835 ton [mi.sup.-2] [yr.sup.-]) for L1, L2, L3, and L4, respectively. The same scenarios caused increases in water erosion from 1125 t [km.sup.-2] [yr.sup.-1] (2866 ton [mi.sup.-2] [yr.sup.-1]) for L0 to 1168, 1332, 1350, and 1677 t [km.sup.-2] [yr.sup.-1] (2976, 3394, 3440, and 4273 ton [m i.sup.-2] [yr.sup.-1]). Hence, converting grasslands to dry crop fields caused more water and wind erosion than increased temperature.

Statistical regression analysis was performed to summarize simulation results and to investigate the possible effects of interactions between temperature and land use

[E.sub.wind] = [[[sigma].sup.s=4].sup.0] [a.sub.s] [L.sub.s] + B[delta]T + [[[sigma].sup.s=4].sup.0] [b.sub.s] [L.sub.s] [delta] T (8)

[E.sub.water] = [[[sigma].sup.s=4].sup.0] [c.sub.s] [L.sub.s] + D[delta]T + [E.sub.wind] = [[[sigma].sup.s=4].sup.0] [d.sub.s] [L.sub.s] [delta]T (9)

where

[E.sub.wind] and [E.sub.water] and are simulated average wind and water erosion, respectively, and are regarded as dependent variables of regression. [L.sub.s] for s = 0 to 4, are five independent indicator variables for five land use types. For example, [L.sub.1] = 1 if L1 is the land use scenario, otherwise [L.sub.1] = 0. The same rule applies to other indicator variables.

[delta]T is another independent variable for temperature increase ([delta]T 0, 1,2, 3[degrees]C, or 0, 1.8, 3.6, 5.4[degrees]F).

Parameters [a.sub.s], [b.sub.s], [c.sub.s], [d.sub.s], B and D, were estimated by regression.

No statistically significant interactions were found between temperature and land use. Table 2 lists other coefficients obtained by regression.

Increasing temperature by 1[degrees]C (1.8[degrees]F) increased wind erosion by 31 t [km.sup.-2] [yr.sup.-1] (79 ton [mi.sup.-2] [yr.sup.-1] and decreased water erosion by 5 t [km.sup.-2] [yr.sup.-1] (12.74 ton [mi.sup-2] [yr.sup.-1]). The combined net increase in soil loss with a 1[degrees]C (1.8[degrees]F) increase in temperature was 26 t [km.sup.-2] [yr.sup.-1] (66.25 ton [mi.sup.-2] [yr.sup.-1). We would expect that overall soil erosion would increase in the future with a progressively drier climate. Land use scenarios L1, L2, L3, and L4 increased average wind erosion by 128, 233, 262, and 837 t [km.sup.-2] [yr.sup.-1] (326, 594, 668, and 2133 ton [mi.sup.-2] [yr.sup.-1]) and average water erosion by 43, 206, 224, and 549 t [km.sup.-2] [yr.sup.-1] (110, 525, 571, and 1399 ton [mi.sup.-2] [yr.sup.-1]), respectively. Combined increases in soil loss resulting from changed land use were 170, 439, 486, and 1386 t [km.sup.-2] [yr.sup.-1] (433, 1119, 1238, and 3532 ton [mi.sup.-2] [yr.sup.-1]), or 2.2,5.7,6.3, and 18%, for L1, L2, L3, and L4, respectively. Converting 1 [km.sup.2] (0.386 [mi.sup.2]) of grasslands to dry crop fields would on average increase soil erosion by 0.38 t [km.sup.-2] [yr.sup.-1](0.98 ton [mi.sup.-2] [yr.sup.-1]),

To get a sense of the relative importance of land use versus temperature, we set the incremental increase to 1[degrees]C (1.8[degrees]F), based on 40 years of temperature data (Gao 1999). Similarly, maximum grassland area converted to dry crop fields was set to 498 [km.sup.2] (189 [mi.sup.02]), about equivalent to the total area of dry and irrigated crop fields today. Most of these crop fields were converted from grasslands during the past century. This maximum change in land use would give an increase of 191 t [km.sup.-2] [yr.sup.-1] (487 ton [mi.sup.-2] [yr.sup.-1]) in average soil erosion, compared to an increase of 26 t [km.sup.-2] [yr.sup.-1] (66 ton [mi.sup.-2] [yr.sup.-1]) induced by the 1[degrees]C (1.8[degrees]F) increase in temperature. Converting grasslands to dry crop fields was about eight times as detrimental to local soil conservation as increased temperature. Hence, our simulation result suggests that human--induced land use change, rather than climate change, is a major factor responsible fo r desertification in northern China.

Summary and Conclusion

A semi-empirical model of soil erosion was developed to simulate wind and water erosion under climate and land use changes. The model adequately simulated annual erosion under current climate and land use scenarios. Hence, the model was partially validated and justified for this case study.

Simulations for climate and land use changes indicated that increasing temperature by 1[degrees]C (1.8[degrees]F) increased annual average wind erosion by 31 t [km.sup.-2] [yr.sup.-1] (79 ton [mi.sup.-2] [yr.sup.-1]) and decreased annual average water erosion by 5 [km.sup.-2] [yr.sup.-1] (12.74 ton [mi.sup.-2] [yr.sup.-1]), If global warming is the main trend of future climate change, the effect of future climate on overall soil conservation would be negative. As a result, more conservation effort is need to offset the additional erosion caused by increased temperature. Converting grasslands into dry crop fields increased wind and water erosion. Specifically, converting 1 [km.sup.2] of grasslands into dry crop fields increased average annual soil erosion by 0.38 t [km.sup.-2] [yr.sup.-1] (0.98 ton [mi.sup.-2] [yr.sup.-1]). No statistically significant interaction existed between land use and temperature. However, changes from grasslands to dry crop fields were about eight times as detrimental to soil conserv ation as incremental changes in temperature. Our results also suggest that land use changes during the past century might be a major causal factor of desertification in sandy areas of northern China. Maintenance and improvement of land use patterns would be effective in combating future desertification.

Acknowledgments

This research was jointly supported by the Natural Science Foundation of China under Grant numbers 39990490, 39725006, and 39899370 and the Chinese Ministry of Science and Technology under grant number G2000018605.

[Figure 2 omitted]

[Figure 5 omitted]
Table 1

Grid cell numbers for soil and land use classes Land use classes

Land
use                             Soil clases
classes       CALC       HSND      MSND      GRAV      WETL

DUNE            1        158        99        2         7
DCRP           247       223        46        54       131
WCRP            4         12        2         3         7
TREE           77        481       301        46       157
SHRB           93        353       272        45        79
GRAS           814       2513      561       762       651
NERO           69        158        30        29        95

TOTAL         1305       3898      1311      941       1127

Land
use          Soil
            clases
classes   OTHR      TOTAL

DUNE       3         270
DCRP       18        719
WCRP       2         30
TREE       17       1079
SHRB       7         849
GRAS      128       5429
NERO      116        497

TOTAL     291       8873
Table 2

Regression of simulated average erosion against temperature and land use
changes using Equations 8 and 9.

Regression          Regression for water
 for wind
  erosion           erosion [E.sub.wind]
[E.sub.wind
     ]

B           30.557              D              -5.105
[a.sub.0]   6557.7          [C.sub.0]          1125.9
[a.sub.1]   6685.6          [C.sub.1]          1168.4
[a.sub.2]   6790.5          [C.sub.2]          1332.2
[a.sub.3]   6819.6          [C.sub.3]          1349.5
[a.sub.4]   7395.1          [C.sub.4]          1674.5
[R.sup.2]  0.9999997        [R.sup.2]         0.9999995
s            3.949              s               0.925
p <         0.0000             p <            0.000000


REFERENCES CITED

Abrahams, A.D., A.J. Parsons, and S.H. Luk. 1989. Distribution of depth of overland flow on desert hill-slopes and its implications for modeling soil erosion. Journal of Hydrology 106:177-185.

Bengtson, R.L. and G. Sabbagh. 1990. USLE P factors for subsurface drainage on low slopes in a hot, humid climate journal of Soil and Water Conservation 45:480-482.

Botterweg, P., R. Leek. E. Romstad, and A. Vatn. 1998. The EUROSEM-GRIDSEM modeling system for erosion analyses under different natural and economic conditions. Ecological Modelling 108:115-130.

Brown, L.R. 1984. The global loss of topsoil. Journal of Soil and Water Conservation 39:162-166.

Buckley, R. 1987. The effect of sparse vegetation on the transport of dune sand by wind. Nature 325:428-428.

Dregne, H.E. 1992. Erosion and soil productivity in Asia. Journal of Soil Water Conservation 47:8-13.

Fentie, B., C.W. Rose, K.J. Coughlan, and C.A.A. Ciesiolka. 1997. The role of the geometry and frequency of rectangular rills in the relationship between sediment concentration and stream power. Australian Journal of Soil Research 35:1359-1377.

Fryrear, D.W,A. Saleh, J.D. Bibro, H.M. Schomber, J.E. Stout, and T.M. Zobeck. 1998. Revised Wind Erosion Equation (RWEQ).Technical Bulletin No. 1, U.S. Department of Agriculture, Agricultural Research Service, Lubbock, TX.

Gao, Q. 1999. Progress report on development of Desertification Forcast System (DFS) for Erdos Plateau. Beijing, National Bureau of Combating Desertification.

Hagen, L.J. 1991. A wind erosion prediction system to meet user needs. Journal of Soil and Water Conservation 46:106-111.

Heimlich, R.E. and N.L. Bills. 1984. An improved soil erosion classification for conservation policy. Journal of Soil and Water Conservation 39:261-267.

Higgitt, D.L. 1993. Soil erosion and soil problems. Progress of Physical Geography 17:461-473.

Kinnell, P.I.A. and L.M. Risse. 1998. USLE-M: emperical modeling rainfall erosion through runoff and sediment concentration. Soil Science Society of America Journal 62:1667-1672.

Laflen, J.M., W.J. Elliot, J.R. Simanton, C.S. Hozhey, and K.D. Kohl. 1991. WEPP: soil erodibility experiments for rangeland and cropland soils. Journal of Soil and Water Conservation 46:39-45.

Laguna, A. and J.V. Giraldez. 1993. The description of soil erosion through a kinematic wave model. Journal of Hydrology 145:65-82.

Loch, R.J. 1996. Using rill/interrill comparisons to infer likely responses of erosion to slope length: implications for land management. Australian Journal of Soil Research 34:489-502.

Merrill. S.D.,A.L. Black, D.W. Fryrear, A. Saleh, T.M. Zobeck, A.D. Halvorson, and D.L Tanka. 1999. Soil wind erosion hazard of spring wheat-fallow as affected by long-term climate and tillage. Soil Science Society of America Journal 63:1768-1777.

Panuska, J.C., I.D. Moore, and L.A. Kramer. 1991. Terrain analysis: integration into the agricultural nonpoint source (AGNPS) pollution model. Journal of Soil and Water Conservation 46:59-65.

Parsons, A.J., J. Wainwright, A.D. Abrahams, and J.R. Simanton. 1997. Distributed dynamic modelling of interrill overland flow. Hydrological Processes 11:1833-1860.

Perrens, S.J. and N.A. Trustrum. 1985. Soil erosion assessment for conservation policymaking. Journals of Soil and Water Conservation 40:491-504.

Renard, K.G., G.R. Foster, G.A. Weesies, and J.P. Porter. 1991. RUSLE, revised universal soil loss equation. Journal of Soil and Water Conservation 46:30-33.

Renard, K.G. and JR. Freimund. 1994. Using monthly precipitation data to determine the R-factor in the revised USLE. Journal of Hydrology 157:287-306.

Renard, K.G., G.R. Foster, G.A. Weesies. D.K. McCool, and D.C. Yoder. 1997. Predicting Soil Erosion by Water: A Guide to Conservation Planning with the Revised Universal Soil Loss Equation (RUSLE). U. S. Department of Agriculture, Agriculture Handbook No. 703,404 pp.

Roo, A.P.D., R,J. Offermans, and N.H. Cremers. 1996. LISEM: a single-event, physically based hydrological and soil erosion model for drainage basins. II: sensitivity analysis, validation and application. Hydrological Processes 10:1119-1127.

Roo, A.P.D., C.G. Wesseling, and C.J. Ritsema. 1996. LISEM: a single-event physically based hydrological and soil erosion model for drainage basins. I: theory, input and output. Hydrological Processes 10:1107-1117.

Rosenberg, N.J., B.L. Blad, and S.B. Verma. 1983. Microclimate. New York: John Willey and Sons.

Smith, R.E., D.C. Goodrich, and J.N. Quinton. 1995. Dynamic, distributed simulation of watershed erosion: the KINEROS2 atid EUROSEM models. Journal of Soil and Water Conservation 50:517-520.

Tiscareno-Lopez, M., M.A. Weltz, and V.L. Lopes. 1995. Assessing uncertainties its WEPP's soil erosion predictions on rangelands. Journal of Soil and Water conservation 50:512-516.

Tucker, C.J. 1979. Red and photographic infrared linear combinations for monitoring vegetation. Remote Sensing of Environment 8:127-150.

Wang, J.A., W.Q. Wang, and H.X. Zhao. 1992. Mapping and analysis of soil erosion for Huangpuchuan Watershed. In, Z.P. Jin, P.J. Shi, F.C. Hou, and H.X. Zhao (eds.) System Models of Soil Erosion its Huangpuchuan Watershed. Beijing: Ocean Publisher, Beijing, pp. 36-44.

Wasson, R.J. and P.M. Nanninga. 1986. Estimating wind transport of sand out vegetation surface. Earth Surface Processes and Landforms 11:505-415.

Weltz, M.A., M.R. Kidwell, and H.D. Fox. 1998. Influence of abiotic and biotic factors in measuring and modeling soil erosion of rangelands: state of knowledge. Journal of Range Management 51:482-495.

Williams, J., M. Nearing, A. Nicks, E. Skidmore, C. Valentin, K. King, and R. Savabi. 1996. Using soil erosion models for global change studies. Journal of Soil and Water Conservation 51:381-386.

Wischmeier, W.H. 1976. Use and misuse of the universal soil loss equation. Journal of Soil and Water Conservation 31:5-9.

Wischmeier, W.H. and D.D. Smith. 1965. Predicting rainfall erosion losses from cropland east of the Rocky Mountains. Agricultural Handbook. U. S. Government Printing Office, Washington, D.C.

Wolfe, S.A. and W.G. Nickling. 1993. The protective role of sparse vegetation in wind erosion. Progress its Physical Geography 17:50-68.

Yoder, D. and J. Lown. 1995. The future of RUSLE: inside the new revised universal soil loss equation (special issue: water research and management in semiarid environments). Journal of Soil and Water Conservation 50:484.

Zhang, H. and M. Shao. 1997. Simulation study on dynamic mechanism of land use and soil erosion in loess hilly regions of Shanxi and Shaanxi Provinces. Science Bulletin (China) 42:1883-1890.

Zhang, X., T.A. Quine, and D.E. Walling, 1998. Soil erosion rates on sloping cultivated land on the Loess Plateau usear Ansai, Shaanxi Province, China: an investigating using [Cs.sup.137] and rill measurements. Hydrological Process 12:171-190.

Zhang, X.C., M.A. Nearing, W.I'. Miller, L.D. Norton, and T.L. West. 1998. Modeling interrill sediment delivery. Soil Science Society of America Journal 62:438-444.

Qiong Gao is a professor and Mei Yu is a postdoctoral fellow at the MOE Key Lab of Environmental change and Natural Disaster, Institute of Resources Science, Beijing Normal university, Beijing, china. Longjun Ci is the director of the National Bureau to combat Desertification, state Forestry Administration, china.
COPYRIGHT 2002 Soil & Water Conservation Society
No portion of this article can be reproduced without the express written permission from the copyright holder.
Copyright 2002 Gale, Cengage Learning. All rights reserved.

Article Details
Printer friendly Cite/link Email Feedback
Author:Gao, Q.; Ci, L.; Yu, M.
Publication:Journal of Soil and Water Conservation
Article Type:Abstract
Geographic Code:9CHIN
Date:Jan 1, 2002
Words:6635
Previous Article:Using soil properties to predict forest productivity in Southern Illinois.
Next Article:An investigation of spatial variation in soil erosion, soil properties, and crop production within an agricultural field in Devon, United Kingdom.
Topics:

Terms of use | Privacy policy | Copyright © 2019 Farlex, Inc. | Feedback | For webmasters