Open AccessModelling of silt content using geographically weighted regression
Author(s) -
Henny Pramoedyo,
Sativandi Riza,
Deby Ardianti,
Affiati Oktaviarina
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/1968/1/012029
Subject(s) - linear regression , curvature , regression analysis , silt , digital elevation model , elevation (ballistics) , statistics , regression , weighting , mathematics , soil science , geology , geometry , geomorphology , remote sensing , medicine , radiology
Multiple linear regression is a method used to model or predict an object that sees the relationship between a dependent variable and a group of independent variables. Geographically Weighted Regression (GWR) is the development of multiple linear regression involving geographical factors. In this study, both methods were used in the study to analyze one of the soil elements, namely the silt soil texture. Through the Digital Elevation Model (DEM) data, the topographic variables used in the study are Eastness Aspects (Ae), Northness Aspects (An), Slope (S), Unsphericity Curvature (M), Vertical Curvature (Kv), Horizontal Curvature (Kh), Accumulation Curvature (Ka ) and Elevation (Elv). The results showed that the GWR model with fixed Gaussian weighting better than the multiple linear regression model. R 2 value of GWR was 57 %, greater than the multiple regression, which was 55%. And the SSE of GWR and multiple regression value were 2014,69 and 2177,19 respectively.