Integration of TUSLE in SWAT Model for Sediment Prediction at a small mountainous catchment, Chenyulan Watershed

碩士 === 國立聯合大學 === 土木與防災工程學系碩士班 === 106 === A number of studies indicated that sediment in water will affect the nutrient in stream. Therefore, before the simulation of nutrient, sediment concentration or sediment transport need to be calibrated and validated first. In Taiwan, it has been observed th...

Full description

Bibliographic Details
Main Authors: LU, CHIH-MEI, 盧至美
Other Authors: CHIANG, LI-CHI
Format: Others
Language:zh-TW
Published: 2018
Online Access:http://ndltd.ncl.edu.tw/handle/q9peqc
Description
Summary:碩士 === 國立聯合大學 === 土木與防災工程學系碩士班 === 106 === A number of studies indicated that sediment in water will affect the nutrient in stream. Therefore, before the simulation of nutrient, sediment concentration or sediment transport need to be calibrated and validated first. In Taiwan, it has been observed that increasing rainfall intensity results in more landslides, debris flows, and flooding at mountainous regions. Steep catchments and high geological vulnerability result in difficulties in predicting soil erosion and sediment transportation. In this study, we integrated the Taiwan Universal Soil Loss Equation (TUSLE) and landslide estimating equation in SWAT model (Soil and Water Assessment Tool) as SWAT-T model, which can more reasonably simulate soil loss and sediment transport in the Chenyulan watershed. The stream flow, sediment yield and sediment transport in Chenyulan watershed were calibrated and validated and the performance of 4 sediment equations in SWAT model (Bagnold equation, Kodoatie equation, Molinas and Wu equation, Yang equation) also were examined. The results show that the simulation of stream flow was satisfactory (R2 ≥ 0.70, RSR < 0.70, NSE > 0.50, |PBIAS| < 25%), sediment transport simulation by Bagnold and Yang equation could simulate sediment transport well (R2 ≥ 0.65、RSR ≤ 0.60、NSE > 0.60、|PBIAS| < 10%). Compared to the original SWAT model (SWAT2016), SWAT-T could significantly reduce the overestimated sediment yield by 70.92%, especially in highly steep areas. Therefore, the soil erosion characteristics of Taiwan should be considered for more reasonable sediment simulation.