Land Suitability Analysis Based on Management for Rubber Plantation Using Mamdani’s Fuzzy Inference System

Abstract

Thailand is the world's largest natural rubber producer and exporter. The most important factor in gaining higher productivity is plantation management. This research, therefore, aims to use Mamdani’s fuzzy inference system (FIS) to evaluate land suitability based on rubber plantations in Nakhon Ratchasima and Buriram provinces. The system is comprised by developing fuzzy membership functions of criteria, estimating criterion group indexes of soil physical and chemical properties, climate, and plantation management, establishing and applying fuzzy rules to agglomerating fuzzy classes of indexes and air-dried rubber sheet productivity, and defuzzifying to attain crisp productivity and generate land suitability maps of different management levels. The agglomeration was raster-based analysis using GIS facilities. The study proved that the productivity was strongly dependent on the levels of plantation management. Higher application levels resulted in higher yields. The study was validated by the use of 30 samples, the relationship of modeled results and observed data was about 1:1 and the RMSE was 25.67 kg/rai/year.