Estimation of Soil Erosion and Statistical Analysis of Controlling Factors in Shihmen Reservoir Watershed
Date Issued
2010
Date
2010
Author(s)
Chi, Wen-Ting
Abstract
The estimation of sediment yield is one of the major topic in the watershed management. Because of the lack of information, the disaster from sediments can not effectively be controlled. Therefore, the primary focus of watershed management is to estimate soil erosion rates and sediment yield correctly. In the past, estimation of soil erosion rates is usually based on Universal Soil Loss Equation (USLE), but it is not easy to obtain accurate parameter factors, because of the difficulty of field investigations. 137Cs technique is not restriced in time and space, and it can be used to estimate long-term soil erosion rate and identify the sources of sediment. These characteristics on soil erosion research has great benefits.
It is a difficult work to collect soil samples to analyze 137Cs inventory, and the sampling process is time-consuming in the Shihmen Reservoir Watershed. Based on collected 1,265 137Cs samples (Chiu, 2008), we use regression and the relevance analyses of impact factors, and then establish the equation to estimate soil erosion rate, and compare with ULSE’s results.
In the case of the regression analysis, the cultivated soil samples of Sanguang and Lingiao watershed have a maximum R2 value (R2=0.66). The regression equation incorporated into the cultivated soil samples can also be explained as high as 60% of the variance of soil erosion rate. Therefore, this regression equation has great credibility in the Shihmen Reservoir Watershed.
In this study, soil erosion rate was obviously over-estimated 1.5 times form ULSE than the 137Cs results. The soil erosion rate of uncultivated soil samples (forest and grassland) are over-estimated about 5 times, and 3times of cultivated soil samples (farmland). USLE does not consider the effects of soil deposition, so that the estimation is often over-esimated. However, the estimation of soil erosion rate by 137Cs technique such as forestland: 3 tha-1yr-1, grassland: 3.8 tha-1yr-1, orchard51.9 tha-1yr-1, and farmland: 45 tha-1yr-1 has the same trend with the ones from USLE.
Subjects
Cs-137 technique
soil erosion
regression analysis
SDGs
Type
thesis
File(s)![Thumbnail Image]()
Loading...
Name
ntu-99-R96521303-1.pdf
Size
23.32 KB
Format
Adobe PDF
Checksum
(MD5):68cf81694395c12aec74a6cdd2f32dca