Rice Security Assessment Using Geo-Spatial Analysis

Abstract

Rice is one of the most important food to half world’s population. The agricultural administrative officer has strategies to change rice into other crops, while rice production was significantly decreased. This study aimed on identifying paddy rice field based on phenology in Landsat 5 of 2009 and Sentinel-2 images in 2017 and assessing rice security using spatial analysis techniques. Object-Based Image Analysis (OBIA) and hierarchy classification methods were applied to identify field by Normalize Different Vegetation Index (NDVI). Suitability area of rice growth with Analytic Hierarchy Process (AHP) statistical model was overlaid by soil properties, meteorology, distance from the water and road factors. The suitable areas and rice field were used to estimate production. Rice security was evaluated using rice yield and rice consumption. Rice maps were overall accuracy of 90 and 96%, kappa accuracy of 0.81 with 0.92. The spatial and spectral information of Sentinel-2 were better to demonstrate field objects than using Landsat 5 data and phenology feature of rice based on NDVI which were highly efficient to extract the field. The factors had higher weight of soil drainage, average rainfall and soil pH respectively, and higher rice growth area showed moderately more suitable than other classes. Yield estimation illustrated less percent of error, and this approach was agreed with physical properties in large area to define levels of rice growth. Rice security was the most stable in Kosum Phisai and Borabu districts in 2009 and 2017, and there was high quality of soil and water resources located in highly and moderately suitable for rice growing in current. The related agencies can apply the results to manage and plan rice cultivation.