Studies on prediction of the eruption of Nisi-no Sima volcano-Ⅱ.
As a part of the national programme on the prediction of volcanic eruption, the Hydrographic Departmentis conducting the investigation on the remote sensing technique to be applied to the submarine volcanoes. Among various features relating submarine volcanoes,of particular interest is the correlation between the appearance of discoloured water on sea surface and the volcanic activity beneath it. Hence, by measuring the state of discoloured water, we may predict growth or decay of the volcanic activity. Nisi-no Sima Sinto was born in 1973 with heavy discoloured water in the immediate southeast of Nisi-no Sima (27°15'N, 139°53' E). Since the sea area around these islands can be regarded as one of the best sites for the experiment of remote sensing technique, surveillance has been made by using of the air-borne multiband camera, M K-1, manufactured by I2S Co. Ltd. and the thermal camera manufactured by Nikon Co. Ltd. The first report (Sugiura and Tsuchide, 1977) dealt with the analogue technique with imageries taken by the multiband camera. In the present paper the preliminary experiments on the digital processing is described. Four sets of multiband (B, G, R, IR) imageries have been adopted as sample material. They were taken on 1974 October 29, 1975 August 13, 1976 August 9 and 1976 December 16 (Figs. 2a, 3a, 4a, 5a). Along the coasts of the islands, 25 points (Fig. 6) have been selected to be measured their imagery densities in four colours respectively by vidicon tube scanning. In order to correct the effect of roughness of the sea surface, measurements have been made at 49 ( =7×7) spots within ±0.7 mm from each object point and averaged. The sunspot effect has been corrected through subtraction of density values of the imagery, which were taken on the open sea immediately before or after the exposure on the islands, from those on the object imagery (as is illustrated by example in Table 4). The density values thus corrected (Figs. 2b, 3b, 4b, 5b) indicate some conspicuous colour feature but are still insufficient to express the degree of colouring quantitatively. Then, three kinds of digital experiment have been made by making use of the corrected density values. ( i ) The frames taken on 1976 August 9 and December 16 have the overlapping of 80% with respective neighbouring frames. After the sunspot correction, density values at each object point are compared with those on the neighbouring frames. Most of the differences thus derived (Fig. 8) deviate from 0 more or less, suggesting the incompleteness of the sunspot correction. (ii) Since the imagery on IR-band contains merely the information on physical state between the sea surface and the camera, like waves and clouds, the difference in density values between the imagery on IR-band and those on the other bands may indicate the existence of the discoloured water and its magnitude, excluding the effects due to the physical state as well as to the vignetting. However, the actual differences thus derived (Fig. 9) show larger scattering than the differences by ( i ) above. At least in the present case, the sunspot effect seems to be more significant than the other effects. (iii) Procedures to derive the differences (i) and (ii) have been combined. Namely, density value differences between neighbouring frames on IR-band have been subtracted from those differences on the other bands. Differences thus obtained (Fig. 10〕deviate from 0 systematically on the average. Through further experimental investigation together with the preparation of appropriate softwares,it is intended to establish a standard procedure of digital processing for measuring the colouring of sea surface quantitatively under various conditions.
| Main Author: | |
|---|---|
| Format: | Journal Contribution biblioteca |
| Language: | Japanese |
| Published: |
1978
|
| Online Access: | http://hdl.handle.net/1834/16317 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|