Modifikasi Analytic Hierarchy Process dengan Reduction untuk Kombinasi Kriteria Paling Berpengaruh Pada Kelengkeng

Abstract

The MCDM method, specifically the Analytic Hierarchy Process (AHP) combined with GIS, is widely used in land suitability analysis. However, using excessive criteria increases the complexity, causes inconsistency in weighting, and can lead to bias, which reduces the accuracy of the mapping. This study proposes a modification of the AHP method through a reduction approach to filter out the most influential criteria in land suitability analysis, focusing on longan plant cultivation in Sentool Plantation. A total of 12 criteria, including elevation, slope, water availability, dry months, temperature, distance to rivers, rainfall, CEC, organic carbon, pH, soil texture, and LULC, were initially analyzed, showing that elevation had the highest weight at 13.34%, with S1 (highly suitable) land covering 518.7957 hectares and S2 (suitable) land covering 3.4258 hectares. These criteria were then grouped into four main categories (topography, water source, soil characteristics, and LULC), and criteria with the lowest weight in each group were reduced if the Kappa coefficient value was above 0.6. The results identified five key criteria; elevation, slope, water availability, CEC, and LULC, which were then reweighted through AHP, with elevation receiving a weight of 28.16%. A comparison of the land suitability maps before and after the reduction against the ground truth showed Kappa values of 0.00 and 0.704, respectively, with the 0.704 Kappa value indicating a substantial level of accuracy. These findings confirm that the reduction approach in the modified AHP method can significantly improve the accuracy of land suitability mapping for optimal longan cultivation.