農業損失空間分析與政策涵義：以雲嘉南淹水潛勢地區為例

氣候變遷趨勢下的極端降雨事件對農業有深遠影響，而往往導致農業生產受到嚴重的損害。然而，不同降雨情境下回到農地單元的農業損失空間分布，及其空間分布存在有何種空間異質性，仍未被相關研究清楚指認。因此，本文以空間資料庫與地理資訊系統為基礎，整合農業普查原始資料、農地利用現況、降雨淹水潛勢等相關資訊，以台灣農業重要產區之雲林、嘉義及臺南為對象，探討在不同淹水潛勢的降雨情境下，當一次極端降雨事件發生時，潛在農損金額的空間分布及其空間異質性。研究結果顯示，降雨情境對各地區影響各異，而各類作物產生之經濟損失確實存在空間區位差異，因此需要擬訂不同的災害救助措施以因應各種農損情況，建議的措施包括：農業保險、各類作物生產專區或設施農業等。

關鍵字

潛在農業損失；農業普查；淹水潛勢；地理資訊系統；空間異質性

並列摘要

Since agricultural production is highly correlated with climate and environment conditions, it is anticipated that climate change will have significant impacts on agricultural development. Extreme rainfall is one of the most common phenomenon of climate change, and often results in damage to agriculture. However, spatially explicit assessment of potential damage to agricultural production is still absent, especially that based on units of agricultural land. Yunlin, Chiayi, and Tainan are the major agricultural production regions in Taiwan. Therefore, a geographic information system (GIS), agricultural census data, agricultural land-use survey data, potential flood maps, and other GIS maps are applied in this research to investigate the spatial distribution and spatial heterogeneity of potential damage due to extreme rainfall. Our findings indicate that potential damage resulting from severe rainfall differs by crop as well as production region. Therefore, disaster relief schemes need to be tailored to cope with the various damage circumstances. Applications of agricultural disaster insurance, special production zones for various crops, and strengthening agricultural facilities are suggested.