The principles and degrees to which land use change and climate change affect direct runoff generation are distinctive. In this paper, based on the MODIS data of land use in 1992 and 2003, the impacts of land use and climate change are explored using the Soil Conservation Service Curve Number (SCS-CN) method under two defined scenarios. In the first scenario, the precipitation is assumed to be constant, and thus the consequence of land use change could be evaluated. In the second scenario, the condition of land use is assumed to be constant, so the influence only induced by climate change could be assessed. Combining the conclusions of two scenarios, the effects of land use and climate change on direct runoff volume can be separated. At last, it is concluded: for the study basin, the land use types which have the greatest effect on direct runoff generation are agricultural land and water body. For the big sub basins, the effect of land use change is generally larger than that of climate change; for middle and small sub basins, most of them suffer more from land use change than from climate change.

Precipitation–runoff relationships within a watershed are driven primarily by the interaction of the interplay factors such as climate, land cover and soils (Mariano et al. 2000). As the rapid development of economy, industry and agriculture has taken place in China, the existing pattern of land use actually has changed dramatically, especially since 1990 (Shi et al. 2007). Land use change induced by intensive human activities like urbanization, irrigation and deforestation has obviously affected hydrology, water quality and hydrologic amenities. In terms of hydrologic response, there are mainly two interrelated but separable effects on a watershed: changes in direct runoff volume and changes in peak flow characteristics (Leopold 1968). For example, direct runoff volume would increase as a result of increase of imperviousness because of less soil moisture replenishment and less ground water storage. And another result of increasing imperviousness is the alteration of water retention times in wetlands and lakes and thus increasing flood peaks during storms (Robert & Raymond 2004). Besides, deforestation and reduction of water bodies would oppositely cause the decreasing yield of direct runoff. Basically, the volume of direct runoff is governed by infiltration characteristics related to vegetation cover, topography and soil type. The purpose of this paper is to explore the effect of land use variance on hydrology in terms of direct runoff based on an eight-class land use system in the watershed of Wei River. To date, numerous studies have linked hydrology with remote sensing data to explain some change phenomena on hydrology (Cermak et al. 1979; Sharma et al. 2001). In this paper, the land use maps are derived from 1-km resolution MODIS data in the years of 1992 and 2003 and laid out in GIS system. To assess the effect of land use change on direct runoff volume, the Soil Conservation Service Curve Number (SCS-CN) method, which has been applied on a wide range of catchments in the United States and across the world, is employed (Boughton 1989; Michel et al. 2005; Durbude et al. 2011; Grimaldi et al. 2013). The direct runoff volume can be calculated from a simple formula with parameters only related to curve number (CN). Since there are always more than one CN in the study area, CN was considered to make a weighted average in order to get only one CN value before using the SCS-CN method (Sharma & Singh 1992). But at present, most studies use the weighted precipitation instead of the weighted CN to calculate the direct runoff volume (Symeonakis et al. 2007). The former way does not take the spatial distribution of land use into consideration, while the latter way neglects the spatial distribution of precipitation. In this paper, we divide the study area into several sub basins based on the locations of precipitation gauges. Subsequently, the direct runoff of each sub basin can be computed separately and at last merged together with the weight of the area ratio of sub basin to the whole basin to get the direct runoff for the whole basin.

THE CN METHOD

Basic concept

The CN method originally developed by the SCS (SCS 1956, 1964, 1971, 1993) is one of the most commonly used and efficient methods for determining the approximate volume of direct runoff for a given precipitation event in a particular area. It accounts for many basin characteristics, viz., land use/treatment, hydrologic soil group and antecedent moisture conditions (AMCs) (Hawkins 1993; Ponce & Hawkins 1996; Sahu et al. 2007). The basic assumption of the SCS-CN method is that the ratio of actual soil retention to potential maximum retention is equal to the ratio of direct runoff to potential maximum runoff, which is precipitation minus initial abstraction. After algebraic manipulation and inclusion of simplifying assumptions, direct runoff is calculated by the following expression 
formula
1
where Q is direct runoff depth (mm); P is precipitation depth (mm); S is the potential maximum soil moisture retention (mm); Ia is the initial abstraction (mm), or the depth of water before runoff such as infiltration or vegetation interception. The parameter S is expressed as 
formula
2
where CN is the curve number and ranges from 0 to 100; it is determined by land use, hydrologic soil group and AMC (USDA-SCS 1972).

Factors determining the CN value

The CN represents an empirical relationship between land use, hydrologic soil group and AMC. Therefore, to identify the specific value of CN, all these factors are necessarily taken into consideration.

Land use

Based on CN values estimated for land use types of LANDSAT imagery (Cermak et al. 1979), the CN of eight land use categories with different hydrologic soil types under average AMC are listed in Table 1. In the table, the order of the CN value according to each land use type can be concluded: Water > Urban > Barren area > Agriculture > Wetland > Forest > Grass > Mixture.

Table 1

CN summary with associated land use categories

   Hydrological soil group
 
Land use category Title Imperviousness (%) 
Forest 30 58 71 78 
Mixture 35 60 72 79 
Grassland/shrub 40 63 75 81 
Agricultural area 62 75 83 84 
Wetland 78 78 78 78 
Barren area 72 77 86 91 94 
Urban and built-up 90 89 92 94 95 
Water 100 100 100 100 
   Hydrological soil group
 
Land use category Title Imperviousness (%) 
Forest 30 58 71 78 
Mixture 35 60 72 79 
Grassland/shrub 40 63 75 81 
Agricultural area 62 75 83 84 
Wetland 78 78 78 78 
Barren area 72 77 86 91 94 
Urban and built-up 90 89 92 94 95 
Water 100 100 100 100 

Hydrological soil group

Soils are classified into four hydrologic soil groups as one element for determining CNs (Brakensiek & Rawls 1983). Group A soils have lowest runoff potential and Group D soils have highest runoff potential.

Antecedent moisture condition

The soil moisture condition before runoff generation is another important factor influencing CN value. Based on the 30-day antecedent precipitation for humid area, the soil moisture condition in this paper is classified into three AMC categories: AMC I (<40 mm); AMC II (40–100 mm); AMC III (>100 mm). is assigned to be 0.3 S for AMC I or 0.1 S in case of both AMC II and AMC III (Sharma et al. 2001).

STUDY AREA AND DATA SOURCES

The study area is the basin above Huaxian hydrological station in Wei River Watershed as shown in Figure 1. It is located in the semi-arid and semi-humid climate zone, which belongs to the continental monsoon climate zone. It covers an area of up to 106,289 km2, and has 22 climate stations around. Based on the locations of 22 climate stations, the Thiessen Polygon method is used to divide the entire basin into 22 sub basins.

Figure 1

The study area.

Figure 1

The study area.

One-kilometer MODIS land use images in 1992 and 2003 (Figure 2) and 100-m soil image data provided by the Food and Agriculture Organization of the United Nations are used in this research. According to the classification of soil hydrology group, the whole basin belongs to Class B. Because land use changes little in a short time, it can be assumed that the land use of 1992 can reflect the land use during the period of 1988–1996, which is denoted as Period I, and the land use of 2003 can reflect the land use during the period of 1999–2007, which is denoted as Period II. The daily precipitation data of 22 climate stations during Periods I and II are used in this research, as well as the daily runoff of Huaxian hydrological station in the corresponding period.

Figure 2

Land use map of 1992 (a) and 2003 (b).

Figure 2

Land use map of 1992 (a) and 2003 (b).

RESULTS

Changes in land use

According to the statistical results of the 1992 and 2003 land use maps, the land use changes during 1992 and 2003 are shown in Table 2. From Table 2, for the entire basin, the main change is that forests, grasslands and mixed land are converted to agricultural land and water reduction is relatively obvious. Because wetlands and barren area occupy a very small proportion of land use, and CN values of forests, meadows and mixed land are close and relatively small, so the analysis of changes focuses on agricultural land, urban and built-up, and water. Among them, the most obvious change of agricultural land mainly occurs in sub basins 1, 7, 12, 13 and 14, mostly with an increasing trend. The most obvious increasing of urban and built-up mainly occurs in sub basin 17. The most obvious decreasing of water mainly occurs in sub basins 2, 11 and 17.

Table 2

Land use in 1992 and 2003

  Ratio (%)
 
Land use type 1992 2003 
Forest 11.73 7.38 
Mixture 34.36 27.44 
Grassland/shrub 37.17 26.76 
Agricultural area 16.28 37.66 
Wetland 0.00 0.00 
Barren area 0.00 0.01 
Urban and built-up 0.22 0.75 
Water 0.24 0.00 
  Ratio (%)
 
Land use type 1992 2003 
Forest 11.73 7.38 
Mixture 34.36 27.44 
Grassland/shrub 37.17 26.76 
Agricultural area 16.28 37.66 
Wetland 0.00 0.00 
Barren area 0.00 0.01 
Urban and built-up 0.22 0.75 
Water 0.24 0.00 

Changes in precipitation

Figure 3 shows that the precipitation in Period II is slightly larger compared with that in Period I. In Period I, the linear regression coefficient of the precipitation is negative, which indicates that the precipitation has a trend of decreasing in this period. During Period II, the linear regression coefficient of the precipitation is positive, indicating that the precipitation slightly increased in that period. In the next step, the precipitation data will combine with land use data in order to calculate the direct runoff, and thus to compare the contribution of precipitation change and land use change with the direct runoff change.

Figure 3

The comparison of the precipitation in Periods I (1988–1996) and II (1999–2007).

Figure 3

The comparison of the precipitation in Periods I (1988–1996) and II (1999–2007).

The validation of the simulation

Three base flow separation methods such as direct separation method, parameter separation method and digital filtering parameters were employed to separate base flow of Huaxian station in both Periods I and II. In Figures 4 and 5, the green-shaded area is the direct runoff range which covers the direct runoff calculated by three methods (the full colour versions of these figures are available online at http://www.iwaponline.com/wst/toc.htm). In Figure 4, the solid line represents the simulated direct runoff by the SCS-CN method under the land use of 1992 and the dashed line represents the simulated direct runoff under the land use of 2003 with the precipitation during the period of 1988–1996. As can be seen from Figure 4, the simulated direct runoff by the SCS-CN method is really close to the one by three base flow separation methods. The relative error of the total direct runoff is 5.75%. In Figure 5, the solid line represents the simulated direct runoff by the SCS-CN method under the land use of 2003 and the dashed line represents the simulated direct runoff under the land use of 1992 with the precipitation during the period of 1999–2007. The relative error of the total direct runoff is 24.8%, which is still within an acceptable range. Therefore, it can be considered that the performance of the direct runoff simulation results using the SCS-CN is good.

Figure 4

The simulation of direct flow during the period of 1988–1996. The full colour version of this figure is available online at http://www.iwaponline.com/wst/toc.htm.

Figure 4

The simulation of direct flow during the period of 1988–1996. The full colour version of this figure is available online at http://www.iwaponline.com/wst/toc.htm.

Figure 5

The simulation of direct flow during the period of 1999–2007. The full colour version of this figure is available online at http://www.iwaponline.com/wst/toc.htm.

Figure 5

The simulation of direct flow during the period of 1999–2007. The full colour version of this figure is available online at http://www.iwaponline.com/wst/toc.htm.

Changes in direct runoff

To compare the changes in direct runoff of Period I and Period II, the ratio of the direct runoff to the precipitation is used as the indicator (Sharma et al. 2001). The ratio which is denoted as the yield is calculated as follows: 
formula
3
where is the total direct runoff and is the total precipitation. To distinguish the contributions of land use change and precipitation change to direct runoff change, two scenarios are assumed. One scenario is to keep the precipitation as the constant, and another is to keep the land use as the constant. All the comparisons are based on the land use and precipitation in Period I. The contribution ratio of land use change to precipitation change is expressed as follows: 
formula
4
where is the contribution of the land use change to the direct runoff change; is the contribution of the precipitation change to the direct runoff change; is the yield change of the direct runoff caused by land use change, which is the difference between the yield under land use in 2003 and the yield under land use in 1992 during Period I; is the yield change of the direct runoff caused by precipitation change, which is the difference between the yield under precipitation in Period II and the yield in Period I . The calculation steps are as follows. Step 1: Assume the precipitation is constant, which is the precipitation in Period I, and then calculate the yields, respectively, under land use in 1992 and in 2003. The result shows that, since the yield changes in most of the sub basins are positive, the yields under land use in 2003 increase compared with the ones in 1992 for most of the sub basins. Only sub basins 3, 4, 11, 20 and 21 have a trend of decrease. Combined with the land use in Figure 3, it can be found that: the reduction in sub basins 3 and 11 is mainly due to the conversion from agricultural land into grassland and the significant decrease of water; the reduction in sub basins 4, 20 and 21 is because of the conversion from grassland to mixture. In addition, the greatest change of the yield has taken place in sub basins 1, 3, 12 and 14. Their common feature is the sharp increase in agricultural land. It is concluded that the most important land use types affecting the yield are agricultural land and water, and the least important is mixture land. Step 2: Assume the land use is constant, which is the land use in 1992, and then calculate the yields respectively under Period I and Period II. As can be seen from the result, the yield changes in most sub basins are positive, which indicates that the yields under the precipitation in Period II increase compared to the ones in Period I for most of the sub basins. Only sub basins 1, 2, 4, 14, 19 and 22 decrease, indicating that the precipitation of six climate stations has a trend of reduction. In comparison of steps 1 and 2, the contribution ratio of land use change to precipitation change can be derived. According to the size of the sub basin, sub basins can be divided into large sub basins, middle and small sub basins. As the areas of small sub basins are too small, the main analysis is done on large and middle sub basins. In Figure 6, for most large sub basins, land use change contribution is greater than precipitation change contribution. As shown in Figure 7, for most middle sub basins, precipitation change contribution is greater than land use change contribution.
Figure 6

The contributions of precipitation change and land use change to the yield in large sub basins.

Figure 6

The contributions of precipitation change and land use change to the yield in large sub basins.

Figure 7

The contributions of precipitation change and land use change to the yield in middle sub basins.

Figure 7

The contributions of precipitation change and land use change to the yield in middle sub basins.

CONCLUSIONS

In this paper, based on land use data of 1992 and 2003 and the precipitation data of 1988–1996 and 1999–2007, the SCS-CN method is used to analyze the change of land use, precipitation and the direct runoff. Two scenarios are established here in order to distinguish the land use change and the precipitation change contribution to the direct runoff change. For the study area, 22 sub basins are divided according to the location of climate stations. Based on the analysis on 22 sub basins, it is concluded: the reduction of the yield of direct runoff in sub basins is mainly due to the conversion from agricultural land into grassland and the significant decrease of water; the most important land use types affecting the yield are agricultural land and water, and the least important is mixture land; for most large sub basins, the contribution of land use change is greater than that of precipitation changes; for most middle sub basins, precipitation change contribution is greater than land use change contribution.

ACKNOWLEDGEMENTS

This research was financially supported by the Natural Science Foundation of China (grant nos 51190094 and 51079098) and the Postdoctoral Science Foundation of China (2014M550769). This support is greatly appreciated.

REFERENCES

REFERENCES
Boughton
W. C.
1989
A review of the USDA SCS curve number method
.
Australian Journal of Soil Research
27
(
3
),
511
523
.
Cermak
R. J.
Fieldman
A. D.
Webb
R. P.
1979
Hydrologic Land Use Classification Using LANDSAT
.
Technical Paper no. 67, USAGE
,
Hydrologic Engineering Center
,
Washington, DC, USA
.
Durbude
D. G.
Jain
M. K.
Mishra
S. K.
2011
Long-term hydrologic simulation using SCS-CN-based improved soil moisture accounting procedure
.
Hydrological Processes
25
(
4
),
561
579
.
Hawkins
R. H.
1993
Asymptotic determinations of runoff curve numbers from data
.
Journal of Irrigation and Drainage Engineering
119
,
334
345
.
Leopold
L. B.
1968
Hydrology for urban land planning: a guide book on the hydrologic effects of urban land use
.
US Geological Survey Circular
554
,
18
.
Mariano
H.
Scott
N. M.
David
C. G.
2000
Modeling runoff response to land cover and precipitation spatial variability in semi-arid watersheds
.
Environmental Monitoring and Assessment
64
,
285
298
.
Michel
C.
Vazken
A.
Perrin
C.
2005
Soil conservation service curve number method: how to mend a wrong soil moisture accounting procedure
.
Water Resources Research
41
,
pW02011
,
[PubMed]
Ponce
V. M.
Hawkins
R. H.
1996
Runoff curve number: has it reached maturity?
Journal of Hydraulic Engineering
1
(
1
),
11
19
.
Robert
C. W.
Raymond
K. T.
2004
Effect of changing forest and impervious land covers on discharge characteristics of watersheds
.
Environmental Management
34
(
1
),
91
98
.
Sahu
R. K.
Mishra
S. K.
Eldho
T. I.
Jain
M. K.
2007
An advanced soil moisture accounting procedure for SCS curve number method
.
Hydrological Processes
21
,
2872
2881
.
SCS
1956, 1964, 1971, 1993
Hydrology, National Engineering Handbook
,
Supplement A, Section 4, Chapter 10
,
Soil Conservation Service, USDA
,
Washington, DC
.
Sharma
K. D.
Singh
S.
1992
Runoff estimation using landsat thematic mapper data and the SCS model
.
Hydrological Science Journal
37
(
1
),
39
52
.
Sharma
T.
SatyaKiran
P. V.
Singh
T. P.
Trivedi
A. V.
Navalgund
R. R.
2001
Hydrologic response of a watershed to land use changes: a remote sensing and GIS approach
.
International Journal of Remote Sensing
22
(
11
),
2095
2108
.
Symeonakis
E.
Calvo-Cases
A.
Arnau-Rosalen
E.
2007
Land use change and land degradation in southeastern Mediterranean Spain
.
Environmental Management
40
,
80
94
.
USDA-SCS
1972
SCWS National Engineering Handbook
.
Section 4
.
Hydrology, Soil Conservation Service, US Department of Agriculture
,
Washington, DC
.