基于光谱指数建模的沙井子灌区土壤盐分反演  

Soil salinity inversion in the Shajingzi irrigation district based on spectral index modeling

作　　者：谢俊博 王兴鹏[1,2] 何帅 刘洋 忠智博[2,3,4] 李妍[5] 洪国军 XIE Junbo;WANG Xingpeng;HE Shuai;LIU Yang;ZHONG Zhibo;LI Yan;HONG Guojun(College of Water Hydraulic and Architectural Engineering,Tarim University,Aral 843300,Xinjiang,China;Northwest Oasis Water-Saving Agriculture Key Laboratory,Ministry of Agriculture and Rural Affairs,Shihezi 832000,Xinjiang,China;Institute of Water Conservancy and Soil Fertilizer,Xinjiang Academy of Agricultural Sciences,Shihezi 832000,Xinjiang,China;Xinjiang Production&Construction Corps Key Laboratory of Efficient Utilization of Water and Fertilizer,Shihezi 832000,Xinjiang,China;Department of Resource Utilization and Plant Protection,College of Agriculture,Tarim University,Aral 843300,Xinjiang,China;Institute of Regional Development,Jiangxi University of Science and Technology,Nanchang 330200,Jiangxi,China)

机构地区：[1]塔里木大学水利与建筑工程学院,新疆阿拉尔843300 [2]农业农村部西北绿洲节水农业重点实验室,新疆石河子832000 [3]新疆农垦科学院农田水利与土壤肥料研究所,新疆石河子832000 [4]水肥资源高效利用兵团重点实验室,新疆石河子832000 [5]塔里木大学农学院资源利用与植物保护专业,新疆阿拉尔843300 [6]江西科技学院区域发展研究院,江西南昌330200

出　　处：《干旱区地理》2024年第7期1199-1209,共11页Arid Land Geography

基　　金：兵团财政科技计划资助(2021AB009);国家重点研发计划项目(2021YFD1900805)资助。

摘　　要：为了快速准确地获取干旱地区表层土壤盐分信息,以沙井子灌区为研究区,利用地面采集的0~10 cm和10~20 cm深度的土壤盐分数据,以及同步获取的Landsat 9 OLI遥感影像上相应点位的波段反射率值,组合两波段和三波段光谱指数,建立低植被度覆盖下盐渍化监测SDI1、SDI2、SDI3、SDI4模型,并检验4类模型对不同土层深度土壤盐分的反演精度。结果表明:(1)当土层深度为0~10 cm时,4类盐渍化监测模型对土壤盐渍化等级分类精度分别为73.56%、66.35%、43.75%和74.52%;而当土层深度为10~20 cm时,相应的分类精度分别为61.06%、62.50%、66.35%和64.42%,说明灌区内土层最佳反演深度为0~10 cm。(2)三波段光谱指数构建的SDI4模型优于双波段光谱指数构建的其余3种模型,能够有效反演沙井子灌区土壤盐渍化程度。研究结果可为灌区土壤盐渍化治理和防治提供有效的技术参考。This study aims to rapidly and accurately obtain surface soil salinity information in arid regions.To achieve this,Shajingzi irrigation district,Aksu Prefecture,Xinjiang,China was taken as the research area,ground-collected soil salinity data at depths of 0-10 cm and 10-20 cm were used along with corresponding Landsat 9 OLI remote-sensing images to extract band reflectance values.This led to the creation of two-and threeband spectral indices,which resulted in the development of four remote-sensing salt detection indices(SDI1,SDI2,SDI3,and SDI4)under conditions of low vegetation cover.These four indices were then evaluated for their accuracy in estimating soil salinity at different depths.The results showed that:(1)The classification accuracies of the four indices were 73.56%,66.35%,43.75%,and 74.52%,respectively,for a soil depth of 0-10 cm and 61.06%,62.50%,66.35%,and 64.42%,respectively,for a soil depth of 10-20 cm.These findings suggest that the optimal inversion depth for soil layers in the irrigation district is 0-10 cm.(2)Among the four indices,SDI4,which was constructed with a three-band spectrum,outperformed the others,which are two-band spectrum-based.SDI4 effectively estimates the degree of soil salinization in the Shajingzi irrigation district,providing valuable technical references for managing and preventing soil salinization in irrigated areas.

关 键 词：土壤盐渍化 遥感监测 光谱指数 沙井子灌区 

分 类 号：S156.4[农业科学—土壤学]