DSSAT模擬灌溉對臺灣水稻產量影響及風險地圖繪製

Abstract

近年來，隨著氣候變遷加劇，農業為首當其衝、受到影響的產業之一，尤其面臨乾旱逆境時，灌溉水的供給若出現問題，勢必會對糧食產量造成負面影響，除了調整栽種期、加強引水蓄水設施、研發耐旱品種等相關策略外，農民也須依賴更精準的災害預警系統，再加上動態性地調整田間栽培管理方式，在天然災害造成大規模損失前做出因應措施，盡可能降低生產損失。本研究使用DSSAT作物模式模擬水稻第1期作在不同灌溉情境下的生產情境，並將節水栽培所會造成的減產百分比繪製水稻生產之風險地圖，此模擬將結合2019年的歷史氣候重建資料、涵蓋全臺農用地之土壤檔和水稻品種臺農67號之作物基因參數。本研究所要模擬的灌溉情境有3種如下：首先分別針對「自動灌溉」和「無灌溉」2種情境進行模擬，透過比較各生育期所需之水量及風險程度，找出對於灌溉水最為敏感的生育期，接著模擬第3種情境－僅於缺水敏感期進行灌溉。為了探討節水栽培的可行性，本研究將全臺所有的試驗地分成了12個區域，接著透過成對t檢定將各小區在不同灌溉情境下的產量進行比較，研究結果發現中彰投、宜蘭、花東等地區若行節水栽培、僅在敏感期進行灌溉，的確可在不影響產量的前提下，達到節省用水量的成效，其省水量約為37.8%至76.4%。本研究結合程式語言及作物模式，進行更有效率之模擬並建立繪製風險地圖之系統性方法，未來若能將氣候預測資料結合校正過後的作物模式，相信能夠針對氣候變遷對產量造成的影響進行預測，以利決策者制定相關的調適策略。

With the intensification of climate change in recent year, agriculture is one of the industries that are most affected. Especially when suffering from drought stress, if there are problems of supplying irrigation water, it will have a negative impact on grain production. In addition to adjusting the planting period, strengthening water storage facilities and development of drought-tolerant varieties, farmers must also rely on a more accurate disaster warning system, coupled with dynamic adjustment of field management methods. Also, they need to take measures before the disasters cause large-scale losses, so as to minimize the loss of production. This study uses the DSSAT crop model to simulate the production of rice under different irrigation scenarios, and draws the risk map of rice production based on the percentage of yield reduction caused by saving water cultivation. The simulation will integrate Taiwan ReAnalysis Downscaling data of 2019, soil data of agricultural land, the crop genetic parameters of the rice variety Tainung 67. There are three kinds of irrigation scenarios to be simulated in this study as follows: First, simulate the two scenarios of 'automatic irrigation' and 'non-irrigation', and find out the growth stages that are sensitive to irrigation water by comparing the amount of water required and the degree of risk in each growth stage. And then simulate the third scenario – 'irrigation only during the water-scarcity-sensitive period'. In order to explore the feasibility of water-saving cultivation, this study divided all the experimental fields in Taiwan into 12 regions, and then compared the rice yield under different irrigation scenarios through paired t-test. The result shows that if take water-saving cultivation (that means only irrigate in sensitive growth stage) in central Taiwan and eastern Taiwan, it is possible to achieve the goal that saving water consumption without affecting the yield. The water saving rate is about 37.8% to 76.4%. This study combines programming language and crop models to conduct more efficient simulations and establish a systematic method for drawing risk maps. In the future, if we can combine predictive climate data with corrected crop models, it is believed that the we can forecast the impact of climate change on yield. Thus, decision makers can formulate relevant adaptation strategies to help reducing the production loss.