Abstract

Assessing variations in soil wind erosion (SWE) is critical for identifying key change areas and formulating desertification control strategies. Satellite images with an expansive spatial coverage and temporal repeatability make it possible to monitor the process of soil degradation and its consequences such as SWE. This research aims to model SWE in the eastern shoreline of Urmia Lake in the 2005–2017 period through multiple-criteria decision analysis (MCDA). Soil moisture, soil erodibility (SE), soil crust index, number of snow cover days, wind field intensity, and vegetation fraction were determined as critical factors affecting SWE. The analytic hierarchy process (AHP) method was applied to determine the weight of each factor. High SE and poor vegetation were the most important factors in the developed SWE model. The SE was precisely estimated (relative percent deviation (RPD)=2.01) by the support vector regression (SVR) method using Landsat-8 image. The developed SWE estimation method had an overall accuracy of 81%. Most of the eastern shoreline of Urmia Lake Region was classified in the severe SWE class. Results showed a declining erosion intensity trend from central parts with high wind erosion (47% of the region) to northern and southern parts of the region. Increasing the distance from the lake led to an increase in SWE.

摘要

评估土壤风蚀 (SWE) 的变化对于确定关键变化区域和制定荒漠化控制策略至关重要。具有广泛空间覆盖和时间可重复性的卫星图像使监测土壤退化过程及其后果（如 SWE）成为可能。本研究旨在通过多标准决策分析 (MCDA) 对 2005-2017 年期间乌尔米亚湖东岸线的 SWE 进行建模。土壤水分、土壤可蚀性 (SE)、土壤结皮指数、积雪天数、风场强度和植被比例被确定为影响 SWE 的关键因素。应用层次分析法（AHP）确定各因素的权重。高 SE 和贫瘠的植被是开发的 SWE 模型中最重要的因素。使用 Landsat-8 图像通过支持向量回归 (SVR) 方法精确估计 SE（相对百分比偏差 (RPD)=2.01）。开发的 SWE 估计方法的总体准确率为 81%。乌尔米亚湖区东岸大部为重SWE级。结果表明，从风蚀强度较高的中部地区（47% 的地区）到该地区的北部和南部地区，侵蚀强度呈下降趋势。增加与湖泊的距离导致 SWE 增加。结果表明，从风蚀强度较高的中部地区（47% 的地区）到该地区的北部和南部地区，侵蚀强度呈下降趋势。增加与湖泊的距离导致 SWE 增加。结果表明，从风蚀强度较高的中部地区（47% 的地区）到该地区的北部和南部地区，侵蚀强度呈下降趋势。增加与湖泊的距离导致 SWE 增加。