水分收支法模擬土壤水分含量作為灌溉管理依據

Abstract

Efficient management of water could be attained through correct assessment of soil moisture, while irrigation was operated in upland. To attain this goal, a simple simulation model formulated by data of precipitation and pan evaporation was introduced to calculate the balance of water in field. The data required for analysis and validation of the model were adopted from irrigation experiment of spring crop of corn, 1988, located in Hsueh-Chia upland irrigation station in Chia-Nan area.Five surveyed data of soil moisture during initial growth stage (0-26 day after sowing, LAI ≦1), which were included in treatment A. were fitted by Least Squares method to get the parameters essential for this model, i.e., Ka, Kb (moisture exchange parameter between surface and subsoil) and Kc (the crop coefficient), respectively. Then the simulation of soil moisture in mid as well as late-seasons was made by using ka, kb and the crop coefficient curve. Another set of soil data surveyed in treatment B were also simulated for the validation of this model. The results indicated parameters for this model were readily available from surveyed data at initial growth stage of corn，i.e., Ka=- 0.131 ±0.019，Kb= 134.58 ± 3.00, Kc=0.153 ± 0.097. And with this set of parameters, the status of soil moisture in mid and late seasons could be easily predicted through simulation. The correlation equation between observed and simulated data is: Estimated=0.9133 X (Measured) 一 17.298，n=21, R2=0.9354 and for data set of treatment B: Estimated=0.6756 X(Measured) + 42.913，n=25, R:=0.8505. Accurate assessment of soil moisture of upland in Taiwan is easy and ready to be achieved, as long as simple climatology data and parameters from field experiment data were available. Hence, efficient management of irrigation is feasible and can be expected. 旱作灌溉作業進行時倘能正確地估算土壤水份狀態，則有效率的進水份管理工作是可以期待的。針對此一目的擬利用簡單的水分收支方法來推估土壤水份含量，而所立的模式僅需要降量及皿蒸發量兩項氣象調查資料，如此，便可以進行土壤水份支的計算。本研究取嘉南水利會學甲旱灌中心，77年春作玉米灌溉試驗資料，為模式分析與驗噔之材料。利用玉米生育初期（播種後0～26天，LAI≦1），試驗之A處理，土壤水份含量５個時間點的調查資料，以最小平方估計法，求出模式所三個參數，分別是Ka、Kb（下層土壤水份交換項屬的2個參數）及Kc（作物係數，crop coefficient），其次利用Ka、Kb及全生育期之作物係數曲線再進行生育中期、盛期及末期土壤水份之模擬；以及另取B灌溉處理的一組土壤調查資料，進行摸擬驗證。 結果顯示利用玉米生育初期之調查資料可以順利取得模式的參數值；Ka=-0.131±0.019，Kb=134.58±3.00，Kc=0.153±0.097，且利用此一參數值模擬玉米生育中、後期的土壤水份含量；得到調查與推估值間的相關式為Estimated=0.9133× (Measured)-17.298，n=21，R2=0.9354取B灌溉處理的資料進行模擬，所得調查值與推估值的相關式為Estimated=0.6756 × (Measured) + 42.913，n=25，R2=0.8505。 根據本研究之結果可知，在本省旱作土壤水份之推估，僅需簡單的氣象資料及容易由田間試驗取得之參數值，就可以進行，可以有效率的進行灌溉管理之工作。