This paper discussed the effectiveness in case of applying synthetic aperture radar (SAR) data to the liquefaction potential assessment. In this study, the liquefaction potential assessment with SAR data and Geographical Information were carried out by using LF (Latency Factor) Model developed by ourselves in 1991. Two examination cases executed to make the liquefaction potential maps are as follows: CASE-A) Only using Geographical Information (GI). CASE-B) Using SAR data and GI. The summaries of this study are as follows: 1) The causality between topographical information and the backscattering coefficient was suggested by the analysis based on the quantification method type II and III. It was confirmed that the SAR data was very useful for extracting areas in danger of liquefaction through the discriminated results by using LF model. 2) In case of using SAR data and GI (CASE-B), analysis of the last conditions on the test field, such as soil moisture, topography and land cover, could be performed with the liquefaction potential map. 3) Furthermore, by extracting pixels of different results between CASE-A and CASE-B, a subtracted image was proposed for the practical use of the SAR data. This subtracted image was very useful for supporting the judgement whether dangerous or not for liquefaction.
Maximum likelihood method and a neural network approach are the most common supervised classification method used with remote sensing multispectral image data such as Landsat TM data. In these method, training samples from each desired set of classes on the original data are used to estimate the parameters of the particular classifier algorithm. Consequently, these parameters depend on observed season and latitude of the observed area. In this paper, a season and latitude independent analysis method is developed. Information of the original data are separated into a parameter which depends on season and latitude, and parameters which are independent of these conditions by a self-consistent data correction and a normalization. The condition independent parameters are expanded by three principal terms obtained from typical spectral patterns of water, vegetation and soil. The pattern components are available to analysis and to classify remote sensing multispectral image data under free from the observed conditions and also available to compare directly with data observed on the ground using multispectral radiometer.
The seasonal variations of the sea surface temperatures (SSTs) in Kagoshima Bay are investigated using the NOAA/AVHRR data. For the calculational method of SSTs, the Multi-Channel SST (MCSST) algorithm by Sakaida and Kawamura is adopted, which is known to produce the most accurate SSTs around Japan. The main results are i) the SSTs in the inner basin is higher (lower) than those in the outer basin in summer (winter), ii) the variations of SSTs are large in autumn and spring and small in summer and winter, iii) in the outer basin, SSTs at the eastern side are higher (lower) in winter (summer) than those at the western side and iv) the eastern side SSTs in the inner basin tend to be higher than those at the western side. The observational facts suggest that the ocean water flows into the basin through the eastern side. The inflowing ocean water is also seen in several NOAA/AVHRR images.
For studying the pollution by dust in winter season at Tonghua city in China, the measurements of spectral reflectance were carried out to estimate the concentration of dust in the snow deposited on the ground. In this study, we found the characteristics of spectral reflectance of the snow surface contained with natural minerals and anthropogenic dust. The most effective indices of spectral reflectance for estimating the concentration (mg/cm3) of dust in the snow were dependent on reflectance of visible range (B1) which mean wavelength value of 450 nm, 475 nm and 500 nm. The regression equations for the relationship between BI and concentration (Y) of dust in the snow were expressed in the form of logarithmic expression (Y=a+b log (B1)). The correlation coefficients obtained from estimated formula were more than 0.89, and the standard errors of prediction were less than 1.44. It can concluded that the spectral reflecance method is useful for estimating of the concentration of dust contained in the snow deposited on the ground during winter season in the range 0.5-6.0 mg/cm3 at this area.
Using N-LAND database system, the snow cover area in the Tohoku district in Japan is detected monthly in 1991. The N-LAND data system includes all the information of AVHRR/NOAA date and software to detect cloud area. The cloud cover area detection is based on the threshold method using all the AVHHR channel informations. Monthly free cloud covered image is created through synthesizing some images on near time. The detected snow area on the images are validated by the AMeDAS snow-depth data and the error is estimated to be less than 8%. The change of the snow cover area and distributions from Feburary to April, 1991 are investigated in the Tohoku district.
The purpose of the paper is to develop a practical method for detecting land cover changes in a prefectural scale area by comparing sattelite image with a land use map. First, LANDSAT TM image was compressed into three bands with principle component analysis and then classified into land cover categories coresponding to a geographic map by a supervised level slicing method. The classified image was compared with a geographic map of the area in order to detect the change during about twenty years. The results were validated by the use of ground survey data with GPS information. Furthermore, a bird-eye view image of the classification was produced to enhance the results for administrative application.