An atmospheric correction algorithm using a radiative transfer code combined with NOAA/NCEP global data assimilation system (GDAS) products is one of standard algorithms for five thermal infrared channels of Terra/ASTER. In the present article, this GDAS-based algorithm is validated by three approaches using AVHRR channels 4 and 5 data around Japan. In the first approach, the GDAS-based algorithm is validated using in-situ measurements of lake surface temperature on Lake Kasumigaura. The results show that 1) the error of the GDAS-based algorithm for channel 4 is almost within ±0.5 K in winter though it is almost within ±1 K in summer, 2) the error of the GDAS-based algorithm for channel 5 is much larger than that for channel 4, and 3) the GDAS-based algorithm is much more sensitive to cloud-contamination than the split window (SW) algorithm. In the second approach, the GDAS-based algorithm is compared with the SW algorithm along the nadir track of a satellite. As the results, the GDAS-based algorithm encounters a trouble for a localized dry region whose size is almost equal to one scene of ASTER. This is an essential problem for the GDAS-based algorithm because this trouble is most likely caused by the low spatial/temporal resolution of GDAS products. In the third approach, the GDAS-based algorithm is validated using the precipitable water vapor (PWV) derived from GPS wet delay data. First, the PWV derived from the GDAS profile is compared with that derived from the GPS wet delay data. For 850 permanent GPS stations of Japan, the differential RMS of them is 5.54 mm in early September. And then, the GDAS-based algorithm is compared with the SW algorithm for cloud-free sea surfaces adjacent to cloud-free GPS stations. The GDAS-based algorithm demonstrates a possibility that its accuracy is improved by the scaling of the water vapor profile with the GPS PWV from the nearest GPS station. In the final part of this article, future problems are discussed. Especially, the necessity of validation experiments in the Southern Hemisphere with less permanent meteorological stations is emphasized.
Algorithm deriving the upwelling radiation from the top of the spherical atmosphere-ocean system is proposed. The method was developed using a technique similar to that of the plane parallel atmosphere bounded by a heterogeneous surface. The effect of the diffuse radiation from the adjacent region is particularly evaluated. Numerical result indicated an effect of spherical atmosphere on the upwelling radiation if the incident solar zenith angle is over 80 degrees.
C-band backscatter intensities from lake ice in Northern Alaska showed a reversal phenomenon from freezing period to spring thaw period. The backscattering mechanisms of this phenomenon have not been clearly explained yet. In order to understand a mechanism of this backscatter reversal, we conducted field observations from April to May in 1997. This paper describes a summary of field observation data analysis and a derived assumption for the backscattering mechanism for the spring thaw period. The backscattering increase of the grounded ice is caused by a roughness increase of the ice surface, while backscattering decrease of the floating ice is by a presence of water film of the ice surface. By using the IEM surface scattering model, a validity of the assumed backscattering mechanisms is confirmed.
For automatic classification of satellite imagery by unsupervised method, category assignment has been an important but difficult task. In this paper, we propose an automatic assignment method using a vegetation map with seven merged categories. The merged raster image with UTM coordinate system and 30 m spatial resolution is obtained by converting digital vector vegetation data edited by Japan Environmental Agency. Unsupervised cluster image data are compared with the vegetation categories to produce a contingency matrix and an assignment priority table. The contingency matrix is used for evaluation of maximum classification accuracy, and priority table is used to combine a relationship between cluster and category. At the Kesennuma city 500 × 500 pixels study area, the overall accuracy is greater than 0.7.
The measurement of bidirectional reflectance distribution function (BRDF) of ground surface becomes more important according to the advanced use of satellite optical sensor. Recently, many in-situ instruments have been proposed for the BRDF measurement, and the calculation of the land surface radiation transfer has been attempted. The target area scale of in-situ measured BRDF is very different from satellite based BRDF, and the in-situ measured BRDF generally shows spatial instability. To validate the satellite based BRDF, it is better to measure the BRDF at the many points in the ground site. However, most of the previous instruments can measure only a few points during the short time when the sun-geometry is unchanged, because the instruments are heavy and it takes a long time to measure the BRDF. In this study, new BRDF measurement system that has a potential to correct spatial BRDF instability using the spectroradiometer and CCD Digital Camera is proposed, and its capability and limitation are discussed.