GISdevelopment.net ---> AARS ---> ACRS 1990 ---> Poster Session

Soil erosion mapping with Remote Sensing method on The Loess Plateau ....The example of Mizhi county, Shaanxi province

Zhang Fengrong, Lin Pei, Xia Fuchuan and Wang Defu
Beijing Agricultural University, Ministry of Hydrology, China


Abstract
Soil erosion is a very difficult job with normal method in the hilly and gully Loess Plateau. In this research the color infrared aerial photo was selected as the remote sensing data of soil erosion survey. Soil erosion factors were identified from the color infrared aerial photo. According to the different combination of soil erosion factors, the legend of soil erosion intensity was established for the Loess Plateau hilly and gully region. The whole county was mapped by using that legend of soil erosion intensity. In doing so a new method of mapping soil erosion intensity with remote sensing data in the hilly and gully region of the Loess Plateau was developed. It also resolves the soil erosion modulus to reveals the different erosion degree of different geomorphological locations in a basin and the mechanism of erosion. It provides the basis of designing soil erosion control planning for different situations.

The objective of this study is to probe the method and procedure of soil erosion mapping with remote sensing techniques in hilly and gully area of the Loess Plateau. The test region is Mizhi County of Shaanxi Province.

Selecting Remote Sensing data for mapping soil erosion
Due to longterm's erosion and mankind's activities, the topography is rather scrap. So, it is very difficult to apply remote sensing data to soil erosion survey in this region. Selecting of suitable remote sensing data is the key to the survey of soil erosion intensity. The suitability of remote sensing data to soil erosion survey as judged. Two false color TM images (May 17 and Sep. 23, 1987) two false color SPOT images (May 5, 1987 and Nov. 28, 1988) and the color infrared aerial photo (Sept. 21, 1987) were used, All these data are applicable to mapping of various soil erosion types such as water erosion, water-gravity erosion, and wind erosion. Since the metric resolution of TM and SPOT images is relatively low, terraced field, shallow ditches and gully can not be recognized by visual interpretation. The color infrared aerial photo (at scale 1:50,000), because of its higher metric resolution, was chosen as the data for soil erosion intensity survey in this region.

Quantitative analysis of soil erosion intensity in Mizhi County
The soil erosion intensity remote sensing mapping is done is two steps:
  1. Analysis of Soil Erosion Intensity Data with Hydrologic Data in Mizhi County.

    1. The average of soil erosion modulus is about 20,000 ton/km2/year in the county.

    2. It's believed that the sediment-delivery ratio is 1:1 in the Yellow River basin. This ratio was proposed by checking the runoff and sediment data acquired only from the hydrologic gauging stations, which the amount of sediment deposited in silt arresters have not been considered. By the end of 1987, there were totally 1067 silt arresters in the county.

  2. Determining Soil Erosion Intensity with Stereoscopic Photo Pairs of the Silt Arresters

    By measuring the stereoscopic aerial photo pairs with stereoplotter, the sediment yield in silt arresters can be calculated. In this way the erosion area relating the silt arrester can be determined. The number of years of the silt arresters, since it was built, can be investigated. Finally, the modulus of soil erosion in the basin controlled by the silt arrester can be determined. The steps of this stereoplotting of silt arresters are as follows.

    1. Information Acquisition

      1. In order to acquire and representative values, the distribution of silt arresters in the whole county was mapped.

      2. The number of the years of the silt arresters since they were built was determined.

      3. Topographic maps (at the scale 1:10,000) were used to obtin values of elevation and slop etc.

    2. Stereoplotting.

      Silt yields in the silt arresters were measured using HPZ-2 stereotopograph.

    3. Calculating the Sediment Yeild in Silt Arresters

      Many studies showed that the volume of silt arrester correlates highly with the height of silt arresters (H), the average gradient of gully (J) and the average width of the silting surface (B). According to field survey data, correlation coefficients of V-H/2 or V-H/J can be 0.99 and V-B 0.95. To calculate sediment yield in silt arrester the regression equation was used, which was developed by Mr. Fan Rueiyu in 1985.

      V = 3.648*10-5 H2.172 * B0.897/ J1.081

      Field survey data for 50 silt arresters were compared with the corresponding calculated values, which reveled that for 57% of the 50 arresters, the error was within + 3%, and for 98% of them, the error was within + 5%. Considering the hydrologic gauging data are referred to suspended silt, and do not include trapped amount by silt arresters and the sediment yield of silt arrester includes bedload as well as suspended silt, it is reasonable to conclude that the measured value of sediment yield with stereo plotting method is very close to the corresponding hydrologic data.
Methodology of mapping soil erosin intensity with Remote Sensing
Several considerations must be taken when mapping system of soil erosion intensity is set up: (Table 1 and 2).

Boundary between rill-interrill erosion section and gully-gravity erosion section must be portrayed according to the erosion types and characteristics in different geomorphologic locations (parts).

The grades of soil erosion intensity was set up, based on erosion types and characteristics in different sections.

Table 1. Mapping System of Soil Erosion Intensity in the Rill-Interrill Erosion Section

Category I Category II Category III
1. Water Erosion 11. delicate erosion
(<1000T/km2. Year)		 111. hilltop, terraced field
112. gentle surface of hills
113. vegetation cover >90%
12. slight erosion
(1000-2500T/km2. year)		 121.hillside land 5° - 8°
122. vegetation cover 70-80%
13. medial erosion
(2500-5000t/km2. year)		 131. hillside land 8° - 15°
132. vegetation cover 50-70%
14. intensity erosion
(5000-8000T/km2. year)		 141. hillside land 15° - 25°
142. vegetation coxer 30-50%
15. very intensity erosion
(8000-15000T/km2. year)		 151. hillside land 25° - 35°
152. vegetation cover 10-30%
16. severe erosion
(15000-25000T/km2. year)		 161. hillside land > 35°
162. vegetation cover < 10%


Table 2. Mapping System of Soil Erosion Intensity In the Gully-Gravity Erosion Section

Category 1 Category II Category III
2. Water -Gravity 21. delicate erosion
(<1000t/km2, year)		 211. channel, silt arrester land
212. vegetation cover >90%
22. slight erosion
(1000-2500T/km2, year)		 221. gullyside land 5°-8°
222. vegetation cover 70-80%
23. medial erosion
(2500-5000T/km, year)		 231. gullyside land 8°-15°
232. vegetation cover 50-70%
24. intersity erosion
(5000-8000t/km2, year)		 241. gullyside land 15°-25°
242. vegetation cover 30-50%
25. very intensity erosion
(8000-15000T/km2, year)		 251. gullyside land 25°-35°
252. vegetation cover 10-30%
26. severe erosion
15000-25000T/km2, year)		 261. gullyside land>35°
262. vegetation cover<10%
27. very severe erosion
(>25,000T/km, year)		 271. gully density >20/cm


In the rill-interrill erosion section, terraced field slop and length of hillside field, vegetation cover or land use type are the main factors. in terms of different combination of these factors, the grades of soil erosion intensity were established for that. Terraced field are usually build on gentle hillside, showing distinct texture of spiral shell on the aerial photoes. However, they are very rarely seen on the hillside field when slop is steeper than 20 degrees.

Applcation and test of the mapping system of soil erossion internsity
In order to test the mapping system of soil erosion intensity (tables0 1 and 20, application was made in trial area. firstly, six small watersheds controlled by silt arresters respectively were chosen for mapping soil erosion. Then, the sediment yields of different areas which have different erosion internsity were measured and calculated. Fanally, the calculated values of the sedment yields were compared with the data from field survey. the result was shown as follows:

the boundary between rill-interrill erosion section and guly-gravity erosion section was precisely portrayed. The average area percentage of the rill-interrill erosion section to the total area is 40.105; in gully-gravity erosion secton, the percentage is 59.90%. This value is very close to that one acquired from detailed land resources survey.

Texture of gully is seen distinctly on the aerial photoes. The boundary between the rill-interriall erosion section and gully-gravity erosion section was portrayed along the end of v-shap gullies. So, the precision of the boundary was higher than the boundary portrayed on the topographic map at the same scale.

The erosion intensity in gully-gravity erosion is higher than that in rill-interrill erosion section. Trial data from neighbour counties has also confirmed the result.

The interpreting result of soil erosion intensity from color infrared aerial photoes corresponded to the field survey data of silt arresters. so, the above results show that the mapping system of soil erosion intensity reflects the physical reality, and therefore can be applicable in this region. And it can also be used widely in the other regions under the same geomorphologic conditions as those in Mizhi County.