Flash Flooding Area Prediction by GOES-9 Satellite Data

Abstract

In this study, a new geocomputing index called the Rain Exposure Index (REI) is proposed. REI can be used for flash flooding prediction due to heavy rainfall accumulation. The index was constructed by computing data from GOES-9 satellite images using combinations of image processing and functional constructs. Water vapour data from GOES-9 satellites were collected from 1st January - 20th September, 2005. Thresholding techniques and varied thresholding values from 65-80 with 5 point intervals were used. REI forecasting results were compared with Thai Meteorological Department (TMD) results. The results indicated that 65-70 thresholding values provided a higher percentage of correct prediction than 75-80 thresholding values. Comparsion between the REI and TMD predictions, found that the REI gave a better correct warning of an event than TMD forecasting.