Please use this identifier to cite or link to this item:
Toward Modeling Erosion on Unpaved Roads In Mountainous Northern Thailand
|Title:||Toward Modeling Erosion on Unpaved Roads In Mountainous Northern Thailand|
|Authors:||Ziegler, Alan D.|
|Abstract:||Contributions of road networks and unstable agricultural activities to downstream sedimentation, water shortages, and flooding in mainland SE Asia are not easily determined because scientific understanding of runoff and erosion processes operating on roads is limited. This dissertation work, conducted within the Pang Khum Experimental Watershed (PKEW) in northern Thailand, supports that owing to low saturated hydraulic conductivity (Ks < 1 6 mm h-1), Horton overland flow (HOF) generation occurs more frequently on unpaved PBCEW roads than on other watershed surfaces having higher infiltrability (e.g., mean Ks for agricultural surfaces ranges from 130 to 320 mm h-1)- Because of frequent HOF generation, the road system contributes to stream sedimentation throughout the rainy season. The highly compacted (bulk density = 1.45 Mg m-3) PKEW road surface typically underlies a layer of loose material of finite depth. Instantaneous sediment transport (St) on roads varies because the supply of easily transported surface sediment is constantly altered by overland flow events, traffic, road maintenance, and mass wasting events, both during and between storms. As surface material is removed during an overland flow event, normalized S( declines from an initial peak rate of ~ 3 g J-1 to a steady rate of =0.5 g J-1 The mechanical stress associated with vehicle passes during a storm increases the availability of loose material, producing 2-4 fold increases in St, and sediment concentration (Ct) values. Herein, rainfall simulation data, surveys of traffic phenomena, and soil property measurements were used to parameterize the physics-based KINER0S2 model for simulating road runoff and erosion. During model validation, instantaneous discharge was simulated well (root mean squared error (RMSE) = 14%). However, because KINEROS2 equations do not “describe” road erosion processes accurately, St was simulated poorly (RMSE = 51.6%). To improve modeling, a methodology recognizing the dynamic erodibility (DE) of a road surface was introduced. By explicitly simulating removal of a layer of loose material, the DE modeling technique improved prediction of St (RMSE decreased to 35.4 %). Finally, a systematic approach is presented to implement DE modeling on any road surface where baseline erodibility and sediment availability can be quantified.|
|Appears in Collections:||
Ph.D. - Agronomy and Soil Science|
Please email email@example.com if you need this content in ADA-compliant format.
Items in ScholarSpace are protected by copyright, with all rights reserved, unless otherwise indicated.