การใช้ภาพถ่ายดาวเทียมในการตรวจสอบความแห้งแล้งทางการเกษตร กรณีศึกษาลุ่มน้ำชี Utilization of Satellite imageries in investigation of Agricultural drought: Case study in Chi River basin

Abstract

Development of drought forecasting models, has difficulty in insufficient drought ground truth, leading to a lack of confidence in the accuracy of the forecast. This study aims to use satellite images to indicate agricultural drought, in order to be able to use satellite images instead of agricultural drought ground truth the Chi River Basin, frequently experienced drought, was selected as the study area. MODIS satellite images, land use map and drought data from the Department of Disaster Prevention and Mitigation, as well as agricultural productivity data from the Office of Agricultural Economics, between 2010-2019, were analyzed for the agreement between agricultural drought and various variables calculated from satellite images. The study selected "rice" as a representative agricultural crop for analysis because it is the main economic crop of the Chi River Basin. The process of analyzing the consistency between the indices from satellite imagery (NDVI, LST) retrieves NDVI and LST data only for the corresponding study period and area. The high NDVI results (green plants) found a low LST (cooler surfaces). The low NDVI (outdoor/crops) found a high LST (warmer surfaces). The results from analysis indicate between 2010-2019 A total of 10 years in the Chi River Basin provinces that Normalized Difference Vegetation Index (NDVI) and Land Surface Temperature (LST) data every 16 days can reflect agricultural drought. It was found that when NDVI is between 0.24 - 0.47 and LST higher than 29.60 degrees Celsius, it indicates agricultural drought because rice yields cannot meet the standards. The reliability test using data between 2013 also showed that NDVI and LST can represent agricultural drought with an average accuracy of 70%, implying that validity of utilization of Satellite Images in steady of agricultural ground truth for developing models for agricultural drought forecasting.