摘要：

Muchoil-landformodelingashownhathe predictabilityfopographic parameters derivedrom digital elevationodels (DEM)oroilpatial variability is influencedyheelectionf DEM's gridize.histudy investigatesoil-terrainelationships inoastal dunetultiple DEMesolutionso examine ifuchcale-dependence isbiquitous phenomenon even in low-reliefystems withelativelyomogeneousubstrates. Weollected 193oilamplesromheinduoastal dunefield in western Koreandnalyzed 11 physicalndhemicaloil properties.asedn principalomponentnalysis,ouroilttributes (K+, pH,otal P.nd nitrate)hateflect dominant pedogeomorphological processesnhe dune wereelecteds dependent variables inpatialegression. Predictors included vegetationypes, distanceromheoastline,ndevenerrain indices extractedrom various gridizesf DEM (5, 10, 20, 30, 40).esults demonstratedtrongcale-dependence inheodel prediction:patial patternsfheouroil variables wereest predictedy landform parameters extractedrom different DEMesolutions.pecifically, weoundhe lowestkaike's Informationriterionor K+, pH,otal P.nd nitratetcalesf 5, 20, 30,nd 40,espectively.esidualsf eachoilttributecquiredthesecaleslso possessedhe lowest degreefpatialutocorrelation identifiedyoran's I.hecale-dependent predictability inuroil-landscapeodeling implieshatine-scaleopographic information would notlwayseptimalornderstandingoilpatial variability.uture pedogeomorphologicalodelersherefore needoake intoccountheniquepatial extentndehaviorf individualoilttributeso determineptimalpatialesolutionsf DEM. We proposehatcale-dependentoil-terrainelationshipsaye widelypplicable notion inoil geomorphology, notnlyo variouserrestrialuggedillslopes wherehe notionaseen dominantly examinednd developed,utlsooystems with lowopographicuggednessndniformubstrateuchsoastal dunes

展开