In designing land readjustment, we are often confronted with necessity to determine requirement of water on paddy field area.
Aiming at this, two methods are generally used : the inflow-outflow method and the method of measuring water requirement in depth. The latter is more frequently used in Japan. There are four steps in its operation :
(1) Determination of the sum of consumptive use and percolated water in a sampled paddy field by measuring ponding depth at a certain frequency by a hook gauge.
(2) Estimation of water requirement in depth for the whole area from the result of (1)
(3) Estimation of gross requirement of water for the whole area from the result of (2)
(4) Estimation of the water requirement after the execution of land readjustment from the result of (3).
These four steps are not explicitly distinguished from each other, generally speaking. The mean of measured values on sample fields is too often simply regarded as water requirement for the whole area. So the author reconsidered the estimating method of water requirement by statistical technique with particular interest in step 2. The results are as follows :
(a) There is a difficulty in dividing irrigation period into several units and determining water requirement in depth in each devided period starting from daily data for a sample field. An error about±3 mm/day is inevitable, of the method is inappropriate.
(b) An error about±3 mm/day is inevitable in estimating the mean value of water requirement in depth in whole area from sample data.
(c) There is a difficulty in estimating water requirement in future from the data of a given year.
Thus it became clear that the conventional method brings about too big errors to estimate water requirement of paddy field area.
