学科分类
/ 1
2 个结果
  • 简介:Informationaboutthespatialdistributionofsoilattributesisindispensableformanylandresourcemanagementapplications;however,theabilityofsoilmapstosupplysuchinformationformodernmodelingtoolsisquestionable.Theobjectivesofthisstudyweretoinvestigatethepossibilityofpredictingsoildepthusingsometerrainattributesderivedfromdigitalelevationmodels(DEMs)withgeographicinformationsystems(GIS)andtosuggestanapproachtopredictothersoilattributes.Soildepthwasdeterminedat652fieldobservationsovertheAl-MuwaqqarWatershed(70km2)inJordan.Terrainattributesderivedfrom30-mresolutionDEMswereutilizedtopredictsoildepth.Theresultsindicatedthattheuseofmultiplelinearregressionmodelswithinsmallwatershedsubdivisionsenabledthepredictionofsoildepthwithadifferenceof50cmfor77%ofthefieldobservations.Thespatialdistributionofthepredictedsoildepthwasvisuallycoincidedandhadgoodcorrelationswiththespatialdistributionoftheclassesamalgamatingthreeterrainattributes,slopesteepness,slopeshape,andcompoundtopographicindex.Thesesuggestedthatthemodelingofsoil-landscaperelationshipswithinsmallwatershedsubdivisionsusingthethreeterrainattributeswasapromisingapproachtopredictothersoilattributes.

  • 标签: 土壤深度 数据预测 数字地形 视觉 多元线性回归模型 属性空间