Landscape analysis is usually carried out using a clear sky model, though the importance of landscape under the other weather conditions such as rainy, foggy and snowy conditions is started in many articles and reports. Using computer graphics techniques, it is possible to simulate landscape under such conditions. The fidelity, however, of such simulation images of landscape is not validated on the quantitative basis, which causes difficulties in generating high fidelity landscape images from given weather parameters such as rainfall intensity and relative humidity.
The authors develop a method of generating far-range-landscape images from the weather parameters through organizing and modifying theoretical/empirical formulae on atmosphere/light interaction processes etc.“Weather effects”on far-range-landscape images can be estimated and their fidelity is validated on the quantitative basis. The proposed method is especially useful for practical applications, because many of the weather parameters used here are available from usual meteorological observation data.

View full abstract

Remotely sensed data from satellites are very useful for the global measurement and analysis of the Earth environment and meteorogical changes. For instance, the Normalized Difference Vegetation Index (NDVI), calculated from the Advanced Very High Resolution Radiometer (AVHRR) data of NOAA satellites, has been widely used in monitoring the changes in lands covered by vegetation.
However, it is difficult to estimate the exact values and changes of the NDVI because the remotely sensed data are affected by many radiometric factors.
In the case of the elimination of these effects which were presented in the Thematic Mapper (TM) data of LANDSAT satellite, a relative correction method has already been proposed.
In our study, the method has been extended and applied to the AVHRR data, and the availability of this method for observation of vegetation changes has been discussed.

View full abstract

A priori probabilities of landcover categories in the study area improve the landcover classification accuracy, although the probabilities are very difficult to estimate in advance of the analysis. Algorithms for the decomposition of mixels to pure landcover categories were developed to estimate landcover area ratios in mixels. If the study area is supposed to be a very large mixel which contains several landcover categories, some of the decomposition algorithms of mixed data can be applied to the centroid vector of the study area. The area ratios of landcovers in the study area are equal to the a priori probabilities of landcovers.
The algorithm of maximum likelihood estimation was applied to estimate the a priori probabilities of landcovers in the study area in this research. As a result of this research, the estimation algorithm worked well and the a prior probabilities of landcovers in seven small study sites were estimated very well. Moreover, those estimated a priori probabilities of landcovers improved the accuracy of landcover classification in the study sites.

View full abstract

In order to observe the status of earth's nature and human activity, we are able to utilize the sensor data onboard the orbiting satellite platform with high resolution. It is well known that remotely sensed data have quite unique character of quick, repeat, cheap and reliance, which reflect the features of earth's surface. This technique is often utilized for some of exploration stages such as petroleum and metal mining resources.
JERS-1 SAR data were taken over disastered area before and after the Hyogoken Nanbu Earthquake, and generated interferrogram showing crustal deformation pattern. Also, we have derived lineament map from SPOT HRV data to compare with the pattern of interferrogram.
This paper describing the effectiveness of new concept the usage of lineament pattern. Because, extracted plenty of lineaments from satellite image suppose to be more efficient to detect surface movement occurred by crustal deformation.

View full abstract

A big earthquake hit Kobe and North Awaji to kill over 5, 000 people and to damage many constructions. We received many satellite image data and the related information taken around the date of the earthquake. The high resolution data transfer from the data distribution organization to the end-user was so fast within three days after the earthquake at the fastest case, and within about one month mostly. The analytical procedure in the laboratory delayed because of the many tasks for the data preprocessing. We need more than half a day for the output of one scene even to make the simple hard copy for the visual interpretation. The worst problem in images is the invisible ground surface caused by clouds, foggy or humid atmosphere, or microwave noise. Even SAR data had some noisy invisible parts. Usually geologists do not use the image data which include the saturated white noisy stripes on some parts of the scene caused by some artificial source on the ground. Many of the optical sensors like MESSR and SPOT were not so useful in the cloudy condition. JERS-OPS stereoscopic image data can not provide the useful analysis around the active fault. SPOT satellite data frequently used for the initial analysis because of the higher resolution and the quick delivery as the satellite-born data. The dynamic range of SPOT data are not enough to analyze the surface structure. The interferometric data of paired SAR data help the centimeter order analysis of surface topographic change in regional view. TM data provided lower ground resolution but the higher dynamic range is interested in looking at the regional structure of active faults. The best resolution can be taken in the image of INTER COSMOS KV-1000 that provides the detailed interpretation of the road and other artificial constructions or automobiles, though the dynamic range in the mountain region of the image were very low in the value. Comparative study of JERS-SAR and LANDSAT-TM images suggests that the earthquake faults are very clear in the continuity and the contrast in images but other lines and circle-like structures of valley are not much concordant to each other. The discordant of the extracted structure between SAR and TM is caused by the different direction of sunlught and microwave illumination with about 45 degree difference of direction. It is also suggested that the dynamic range of the brightness of images is much more important and useful specification to analyze the structure in images than the higher ground resolution.

View full abstract