Hydrological Evaluation of Flow Diversion Terraces Using Downhill-Slope Calculation Method for High Resolution and Accuracy DEMs

Junyu Qi,Sheng Li,Fan-Rui Meng,Kang Liang,Lichun Wang

doi:10.3390/su10072414

Abstract

Slope is an important topographical parameter in hydrological models. Dramatic changes in slope are produced due to micro variations of topography in flow diversion terrace (FDT) systems. Assessing the effectiveness of FDT on soil conservation yields different results using different slope calculation methods with different resolution and accuracy digital elevation models (DEMs). In this study, the ArcGIS built-in average-neighborhood-slope (ANS) method and a downhill-slope (DHS) method were used to calculate slopes for a small agricultural watershed in New Brunswick, Canada. Six DEMs were used, i.e., 1, 5 and 10 m resolution DEMs generated with conventional photogrammetric techniques, and 1, 5, and 10 m resolution DEMs derived from light detection and ranging (LiDAR) data. Calculated slopes were summarized for the whole watershed, a patch of FDT systems, and grassed flow channels in front of embankments in the patch. Results indicated that the DHS method produced smaller slopes than the ANS method with LiDAR-1m DEM along grasses flow channels. Furthermore, the mean L-factor derived from the DHS method was smaller than that derived from the ANS in terraces and grassed flow channels for the LiDAR-1m DEM. Finally, soil loss estimated from the DHS method using RUSLE was more consistent with measurement for the watershed than that from the ANS method when the LiDAR-1m DEM was used. The DHS method can provide accurate slopes along grassed flow channels and can be used to assess FDT systems using USLE-based models with high resolution and accuracy DEMs.

Highlights

Assessing the effectiveness of a flow diversion terrace (FDT) system is of importance for landowners and agricultural agencies to make decisions
Six digital elevation models (DEMs) were used, i.e., 1, 5 and 10 m resolution DEMs generated with conventional photogrammetric techniques, and 1, 5, and 10 m resolution DEMs derived from light detection and ranging (LiDAR) data
Results indicated that the DHS method produced smaller slopes than the ANS method with LiDAR-1m DEM along grasses flow channels

Summary

Introduction

Assessing the effectiveness of a flow diversion terrace (FDT) system is of importance for landowners and agricultural agencies to make decisions. Benchmarking or paired watersheds, upstream and downstream monitoring, and edge-of-field testing can help determine the efficiency of the FDT [1] When it comes to large agricultural areas the experimental survey is time-consuming and expensive [2,3]. USLE (universal soil loss equation)-based soil loss prediction algorithms are the most widely accepted and utilized method in hydrological models, such as in the Soil and Water Assessment Tool (SWAT) [4] In those USLE-based algorithms soil conservation support practice factor (P-factor) is often required as an input to estimate the effectiveness of FDTs. Many studies have pointed out that large uncertainty was associated with those monitored P-values [5,6,7]. The P-factor values for various FDT implementations are only valid when the size of a field is greater than that of a FDT system

Objectives

Methods

Results