雲嘉南地區水稻災損數據之診斷分析與未來推估

Abstract

自臺灣產業結構變化後，稻作面積逐年下降。面臨頻繁的氣象災害事件與全球暖化衝擊，讓稻米產量受到影響，且氣候變遷影響全球糧食生產與貿易，將會影響臺灣進口糧食的穩定性。 農作物損失資料常結合災害事件之氣象因子（例如：降雨量、風速）或颱風特性（例如：路徑、暴風半徑、中心最大氣壓等），分析兩者之間的關聯。損失資料大多使用以縣市統計資料；氣象資料則引用中央氣象局、水利署等單位所提供的資料。但使用較大的空間範圍探討或分析災害對產業的衝擊，可能會忽略空間差異。因此，在使用農業災害損失數據時，需考量農業種植現地的特性，以分區、作物之方式，才能使分析較準確。 本研究目的為選擇臺灣主要稻米生產區且受災次數較高的地區，使用詳細至各鄉鎮區之水稻災損資料，分析近年災害事件（豪雨或颱風事件）之水稻損失數據與災害之氣象因子（累積降雨、風速）之關係。考量農業種植現地的特性，萃取可用於分析的損失資料後，再依據損失曲線，選擇冪函數進行參數靈敏度分析，以提供未來水稻損失的推估。本研究氣象資料引用臺灣氣候變遷推估與資訊平台（TCCIP）計畫，所產出的1km網格解析度的日降雨資料。此為高空間解析度之網格化降雨資料，更利於分析現地資料的研究。 Since the change of industrial structure in Taiwan, the area of paddy rice has been declining year by year. Faced with frequent meteorological disasters and global warming impacts, rice production is affected. Climate change affects global food production and trade.This will affect the stability of imported grain in Taiwan. Crop loss data are often analyzed in conjunction with meteorological factors (eg, rainfall, wind speed) or typhoon characteristics (eg, path, storm radius, center maximum pressure, etc.) of a hazard event. Most of the loss data are based on county and city statistics. Meteorological data refer to information provided by the Central Meteorological Bureau and the Water Resources Department. But the use of a larger space to explore or analyze the impact of disasters on the industry, may ignore the spatial differences. Therefore, in the use of agricultural disaster loss data, we need to consider the characteristics of agricultural planting site and separate zoning and crops, in order to make the analysis more accurate. The purpose of this study is to select rice damage areas in Taiwan's major rice-producing areas and use the data which is detailed in this region. Besides, this study also analyze the relationship between rice losses and meteorological factors (Cumulative rainfall, wind speed) of disaters (heavy rain or typhoon events) in recent years. Considering the characteristics of agricultural cultivation, the extraction of the loss data available for analysis, and then based on the loss curve, select the power function for parameter sensitivity analysis to provide future estimates of rice loss. The meteorological data of this study use the Taiwan Climate Change Prediction and Information Platform (TCCIP) program to produce daily rainfall data of 1 km grid resolution. This is a high spatial resolution of the grid of rainfall data and more conducive to the analysis of the current data.