應用DSSAT評估乾旱情境下臺灣農地之作物適栽區

Abstract

農業乾旱是危害作物產量的環境風險之一。隨著氣候變遷的問題日益嚴重，農業乾旱風險的提升對於農業而言實屬挑戰。當作物生產面臨乾旱的問題時，轉作可能是調適策略的方法之一。然而，以農業試驗分析乾旱情境之適栽作物會有時空上的侷限性，不易進行長年期且大範圍的評估。本研究之目的在於應用農業技術轉讓決策輔助系統（Decision Support System for Agrotechnology Transfer, DSSAT），藉由不同灌溉模式之下的模擬作物產量來評估乾旱風險。透過自動灌溉和無灌溉兩種灌溉模式，DSSAT可以平行地模擬出作物面對兩種情境的結果。為了分析長年期的生產結果，本研究結合臺灣歷史氣候重建資料 (Taiwan Reanalysis Downscaling data, TReAD)、農地土壤資訊與各作物的栽培管理進行模擬，並利用乾旱情境指數歸納特定土壤取樣點的適栽度，最後利用地圖呈現出適栽度的空間分佈。 研究結果顯示，在自動灌溉情境下，農作物的平均產量確實顯著地高於無灌溉情境，說明無灌溉對於作物產量會有明顯影響；乾旱情境指數除了評估模擬情境的差距之外，也能夠針對產量進行比較，能同時反映出乾旱影響與產量大小的地區變化。此外，利用地圖顯示了特定區域與作物類型在乾旱期間的脆弱性差異，提供了重要的區域性風險評估資料。此工具可以做為農民選擇適合該區域栽種作物種類的參考指數，評估各作物的優勢環境，規劃適應乾旱的栽種調適策略，協助利害關係人設計更具效益與穩定性的農業栽培制度。

Agricultural drought is one of the significant hazards leading to a substantial reduction in crop yields. Under the impact of climate change, the increasingly severe drought issues pose a challenge to agriculture. When crop production faces drought problems, substituting crops may be one of the adaptive strategies. However, using agricultural experiments to analyze suitable crops under drought scenarios has spatial and temporal limitations, making it difficult to conduct long-term and large-scale assessments. Therefore, this research aims to employ the Decision Support System for Agrotechnology Transfer (DSSAT) to evaluate drought risk by simulating crop yields under different irrigation schemes. Through both automatic irrigation and no irrigation schemes, DSSAT can parallelly simulate the crop yields under these two scenarios. To analyze long-term performance, this research integrates Taiwan Reanalysis Downscaling data (TReAD), field soil information, and the cultivation management of various crops for simulation. A drought scenario index is defined to summarize the suitability of specific soil sampling points, and maps are employed to present the spatial distribution of suitability. The results of the research indicate that, under the automatic irrigation scenario, the average yield of crops is significantly higher than that under the no irrigation scenario, demonstrating the significant impact of no irrigation on crop yields. The drought scenario index not only evaluates the differences in simulated scenarios but also allows for a comparison of yields, reflecting changes in regions with both drought impact and yield amount. The use of maps shows the differences in vulnerability of specific areas and crop types during drought periods, providing important regional risk assessment data. The drought scenario index can serve as a reference index for farmers to select suitable crop types for the region, evaluate the favorable environments for each selected crop, plan adaptive strategies for drought-resilient agriculture, and assist policy-makers in designing more efficient and stable agricultural cultivation systems.