Mt. Usu erupted from 7-14 in Aug. 1977. The volcanic ash accumulated several millimeters to 30 centimeters on the agricultural fields around Mt. Usu.
This study is composed a part of application of remoto sensing technique on the agricultural problem. Using aircraft MSS data, we carried out a digital analysis of the agricultural damage and the distribution of ash fall area. Crop identification on the damaged fields was successful with the supervised most likelihood classification method. The average correct performance by training class was 59.3% in the test fields. The results of large area observation test was successful to classify the degree of damaged fields and forests. In this test, it was effective to use the supervised most likelihood classification method and the total classification performance was estimated as 70.3% in the test fields. Consequently, it is obvious that the sufficient survey of agricultural damage resulting from ash fall is possible with the supervised classification. For this purpose, we must select suitable training fields by the ground truth data. Besides we could make the distribution map of ash fall by the processing of “CH.10-CH.7”.
As mentioned above, a digital analysis of the aircraft MSS data is effective for the surveying of field damage and the estimation of eruptive influence.
View full abstract