To provide necessary data to mass balance models in Lake Shinji, Shimane Prefecture, a methodology for the estimation of gross primary production (Pg) at the surface of the lake using SPOT HRV data was proposed and evaluated using satellite/in-situ water quality data sets acquired in 1995-1998. The procedures of the methodology are, 1) the offset values of HRV bandl (XS1) and band2 (XS2) are calculated from the mean and the standard deviation of XS1 and XS2 in the lake area only, 2) the area average of surface chloropyll-a concentration (Chl.a) is calculated from those of HRV data after offset removal, 3) the average Pg for the whole lake surface (Pg) at the satellite observation time is estimated using the estimated Chl.a from SPOT HRV data. Rough estimations of Pgs based on the proposed methodology were within the range of about 50-700mgC/m3/hr. It was found that Pg in the summer and the autumn was usually 6 to 10 times higher than those in the spring, and Pg under algal bloom condition was twice higher than those of the usual condition even in the same season.
A digital elevation model (DEM) with same map projection and spatial resolution as satellite imagery is required for the correction of topographic effect on radiometric and geometric distortion. As the DEM is used to calculate solar incident angles, the spatial derivative of elevation such as slope is an important feature as well as elevation itself. In this paper, we evaluate interpolation methods for resampling the DEM published by Geographical Survey Institute of Japan into UTM coordinate system. From the practical standpoint, we investigate the simple interpolation methods : nearest neighbor, bilinear and cubic convolution. The converted DEM is compared to the original DEM with respect to statistics of slope angle in addition to elevation. It is shown that the cubic convolution is adequate for the resampling because its parameter can control the smoothness and sharpness of the converted image. However, if the finer resolution is required, the bilinear interpolation seems more adequate than cubic convolution.
The use of synthetic aperture radar (SAR) data for savanna vegetation mapping in central Brazil is investigated. A wet season (February, 1996), L-band, JERS-1 SAR scene over diverse savanna vegetation types in central Brazil was acquired. A statistical analysis of 1133 backscatter samples extracted from the most representative land cover classes of the study area as well as the analysis of the relationship between field leaf area index measurements and corresponding radar returns were the basis of this study. The SAR data separated the grassland, mixed grass/shrub/ woodland and woodland units in the study area. We also found an exponential relationship (r2=0.63) between σ°and LAI. These results indicate the potential of the L-band SAR data for savanna-like vegetation mapping and monitoring.
This paper is concerned with evaluating the compositing algorithms for mosaic NOAA/AVHRR images of Asian region. Five algorithms, such as (1) maximum NDVI method (MaN) by Holben, (2) maximum brightness temperature method (MaT) by Chilar et al., (3) maximum NDVI followed by minimum scan angle method (MaNiS) by Chilar et al., (4) multiple-object composite method (MOC) by Stoms et al., and (5) maximum brightness temperature followed by maximum NDVI and minimum scan angle method (MaTNiS) by the authors, were applied to compose 10-day mosaic images of four seasons by using the data received at Bangkok and Ulaanbaartar. Evaluation was undertaken from the view points of cloud removal, percentage of near-nadir observation data, image smoothness and providing the appropriate data for vegetation growing change. Each algorithm was accompanied with various advantages and disadvantages, but MaTNiS was nominated to be superior as overall.
Adjacency effects from layered box shaped clouds are clarified by means of Monte Carlo Simulation. Effects of optical depth of the clouds and the atmosphere as well as of surface reflectance of the Earth and the clouds are also found together with the effects from the top and the bottom clouds. In particular, effects of cloud bottom height and cloud side planes are also clarified.
The dynamical characteristics of the ASTER Ground Data System (ASTER GDS) were analyzed for the various operational conditions. Dynamical system resource allocation mechanizm, non-linearity of data prodessing capability of multiple data transfer and data output, and user requests for higher-level data products were modelled so as to simulate realistic situatiuons. A cloud-free observation was modeled to check response of the heavy-duty operations. The concluding remarks are as follows ; 1) non-lineality has relativly large impact, 2) however the mutiple data transfer has little impact on the system, and 3) the influence of data output is significant. It is also concluded that the performance of this system is sufficient both in normal operation and heavy-duty operation under a typical assumption of the system parameters. Finally, the studies presented in a series of papers demonstrated the effectivness of system dynamical analysis for the performance analysis of the satellite data processing facilities.
Fronts of water vapor around Bo Hai strait are found by NOAA/AVHRR data. They are detected from January to March and unobserved in other seasons. If the fronts at 1998 Mar. 2nd 3 : 10 and 7 : 05 are identical, it goes from the north west of Huang Hai sea into Bo Hai sea with the velocity of 27 km/hour in the rough.