本論文依據現地水田坵塊系統於稻作期間的用水型態，撰寫一邏輯上合理的演算模式以模擬此期間系統內各單一坵塊田間的動態湛水深。 模式以田間蒸發散量、滲漏量、引水量、排水量、降雨量、側向滲漏量、坵塊空間型態計算田間的動態湛水深變化。水田坵塊系統的回歸水量估算即由模擬歷程中的堰口流出量及田埂側滲量累加而得。 論文中以2000年稻作期間的氣象資料模擬不同型態配置的虛擬水田坵塊系統，發現水稻田與旱作交錯的形式導致水稻田需水量及用水量大幅增加；水田蓮花田交錯的配置形式，僅就水稻田部分的用水量是略為降低的。而以2000年桃園11-2灌區第1單區為實際模擬區域，在模式預設的參數條件下，發現坵塊系統有效暗回歸水量約為坵塊系統總需水量的4.97%?5.64%，但模式中水力傳導係數的擇定對於模擬結果有相當大的影響。 本模式提供了理想狀況下農業灌溉用水精算的可能，而其中水田坵塊系統產生與利用回歸水的方式與單一坵塊迥異，因此有必要就空間上考量回歸水利用的意義。 The thesis makes a logical algorithmic model to simulate dynamic ponding depth of each paddy field in the objective system, stands on ordinary rules of water supply and demand in the cropping seasons. This algorithmic model calculates dynamic ponding depth by the amount of evapotranspiration, percolation, irrigation, drainage, rainfall, lateral seepage and the paddy field’s spatial information. The estimation of the quantity of return flow is accumulating the overflow from weirs and the paddy border’s lateral seepages. The model simulates different spatial distributions of virtual paddy field by the climatic data of year 2000. In the case that rice paddy interlaced with upland crops contrasting with the disposition of all rice paddy field results in obvious increment of water requirement (WR) and field irrigation requirement (FIR). In the other case that rice paddy interlaced with lotus field contrasting with the disposition of all rice paddy field results in slightly decrement of FIR. And the model also simulates the actual paddy field at Tao-Yuan 11-2 irrigation region 1st unit by the climatic data of year 2000 under the hypothetic conditions. The results indicate that the effective invisible return flow occupies 4.97% ~ 5.67% of the paddy field system’s WR. However, the choice of hydraulic conductivity influences the results seriously. This model provides one possible way to calculate actuarial FIR under the ideal state, but the spatial meaning should be paid more attention to for the wide gap between the whole system’s and the single plot’s formation and utilization of return flow.
