Utilization of remote sensing data for estimating damage ratio of rice crop : Part6 Examinationof the yield estimation model based on kinds of agricultural weather hazard

摘要

[ABSTRACT] The evaluation with objectivity is required by the policyholder because the estimate of damage ratio of rice crop is depending on the evaluation member’s skill and also it has misgivings about the evaluation staffs aging. In this study, the estimation of the rice yield affected by the storm and flood damage and the blast disease were conducted in Hokkaido and Miyagi prefecture, Japan. The five different models were adopted to estimate the rice yield of each damage. The predictive error of the blast disease derived from both the multiple linear regression and the ridge regression was within the range of value calculated through the external examination by the evaluation of expert staff. The result suggested that, by the creating each yield estimation model for kinds of agricultural weather hazard, the predictive error will be improved.