灌溉率推估模式之建立

Abstract

台灣目前農業用水為用水的大宗，而灌溉用水又佔農業用水之87%，在面對用水不足的壓力下，農業用水成為最大的轉移目標，因此應配合農業用水結構的改變，灌溉用水量需予以合理估算。目前在灌溉用水的估算方面乃採用「灌溉率」的方式計算之，其來源乃是依據工作人員與農民多年工作經驗粗估得知。使用上雖簡便，但近年來因為作物制度之變更，如休耕、混植造成目前水利會現行之灌溉率已非完全適用，因此本研究中將探討灌溉率之影響因子，建立以灌溉率為基礎之農業灌溉用水量推估系統，估算台灣區域性之水田灌溉用水量，作為研擬合理農業用水標準之基礎。 本研究擬以嘉南水利會為研究區域，建立相關資料庫並採用系統模擬的方式模擬出灌溉用水量，並與水利會以往之灌溉率所估算之計畫用水量進行比較，驗證模式對於用水量推估之正確性與否。接著，在模式中考慮了作物制度變更，意即休耕模擬與混植模擬，探討在不同休耕、混植模擬下與灌溉用水量之變化。最後透過模擬所產生的數據資料進行統計迴歸分析，即可推估得與灌溉率相關之迴歸式。 迴歸結果發現對於灌溉率的推求從耕作型態可分為休耕與混植兩個部分；從耕作的期別分為一期作與二期作，因此從所迴歸出來的式子有休耕之一期與二期及混植之一期與二期共四條，式中發現灌溉率受耕作面積、作物係數、溫度三個因素影響最大。意即對於農民而言，在估算灌溉率之前只要確定作物型態是休耕或混植，以及耕作的期作別是一期作或二期作，決定之後再依據當時的耕作面積、作物係數、溫度代入迴歸式即可推估得灌溉率。因此本研究建立推估灌溉率之方法，估算台灣區域性之水田灌溉用水量，提供給水利單位作為用水規劃之參考依據，以減少農業用水在水資源調配上之浪費，並增加水利會在營運管理方面之便利性，作為提升農業用水效率之依據。

Irrigation consumes more than 69% of the total water supplies in Taiwan. A more efficient and effective irrigation demand planning is the basis for sound regional water resource management. The method of irrigation rate is a traditional irrigation demand estimation tool used by most of the irrigation association in Taiwan. The irrigation rate is the area that can be properly irrigated with a unit flow rate. These empirical indices were developed by the field irrigation engineers through years of field irrigation experiences. But these empirical indices may no longer valid from the recent rapid change in cropping patterns in most of the irrigation command areas due to the socio-economical development and Taiwan’s participation into the WTO in 2002. The objective of this study is to establish a estimation model for irrigation rate under changing environments. The command area of the ChiaNan Irrigation Association, the largest one in Taiwan, is selected as the study area. Related data for regional irrigation demand estimation are collected into a database to establish a regional irrigation demand estimating model. This model is validated using the past cropping patterns and irrigation records. This model is then used as a data generator for generating irrigation demand data under different cropping pattern including fallow. These data are then used as the basis for regression analysis to develop algorithm for irrigation rate estimation under different field conditions. The results show that the irrigation rates are mostly correlated to paddy farming area, crop coefficients, and temperature. There are different formulations for first and second crop of paddy. Whether the area other than paddy farming is fallow or is planned as other crops can also affect the irrigation rate estimation. There are consequently four formulations for the study region. The irrigation planner can estimate the irrigation rate by choosing the corresponded formula for their target field situation. This will increase the efficiencies and effectiveness for regional irrigation demand planning under rapidly changing environment.