1990 Volume 34 Pages 607-612
For building database of irrigation intake water, especially for paddy-fields, data structure must be made clear. The authors collected printed or manuscript irrigation data of 326 points in Kinu and Kokai River basins. 142 points of them have some measured data. By the analysis of these data, following points come to be clear. 1) Irrigation facilities are historical structure. Therefore random additional information cannot be negligible. Database must have some variable length data structure. 2) Some structures have multi-names. Synonym dictionary must be prepared.