This paper discusses a method to make a good use of airborne MSS data as input data for simulating cold currents formed within a valley with forest during the summer nighttime. A suburban area surrounded by Satoyama (urban-neighboring hills and forest, suburban forest) was selected for analysis. The analysis method is as follows. The first step is to classify actual conditions of land cover using the MSS data and GIS data. The second is to analyze the relationship between topographic features and surface temperature distributions at night. From the results of the above analysis, a valley where a residential area is located and the nocturnal cold current occurs was selected for CFD simulation. The 3D model for CFD simulation was created based on the MSS data and GIS data. The surface temperatures from the MSS data were used in CFD simulation as input data. As a result, it was found that the cold current developed above the valley slopes flows down to the bottom of the valley, and the air temperature distribution can be quantified. In order to easily understand the simulated results, the clod current distribution can be visualized in a 3D color image.
Horizontal geometric error is one of the important issues for a comparison between a satellite image and point level observations in plotscale ecological studies. This is because a sub-meter or less than 10-meter horizontal error is required for such investigation. It is recognized that orthorectification with topographical information is an effective method to provide an image with fine horizontal accuracy. Most of commercial high resolution satellites publish an ortho-photo product and its nominal horizontal errors. Although tree height also affects the horizontal errors as well as topographical elevation, orthorectification accuracy in a forest area does not evaluate quantitatively due to difficulties of GCPs on a forest canopy plane. This paper describes the quantitative evaluation of horizontal accuracy of ortho-QuickBird images using the DEMs (digital elevation model), the DSM (digital surface model) and the control points on the top of the tree canopy.
An ocean wind retrival method with microwave radiometer data is proposed taking into accout relative wind directions on observed brightness temperature. The proposed method is based on the geophysical models which was proposed by Frank Wentz and the wind speed retrieval algorithm utilizing a modified simulated anneiling. Validity and usefullness of the proposed method is confirmed with Advanced Microwave Scanning Radiometer (AMSR-E) data and the corresponding NCEP Global Data Assimilation System (GDAS) data. Through a comparison of the wind speed derived from AMSR-E and GDAS, it is found systematic difference which depends on Relative Wind Direction (RWD). Observed brightness temperature from AMSR-E data is modified as to remove the systematic difference which results in improvement of the ocean wind speed retrieval accuracy by up to 61.2% which depends on location and seasons in concern. In the worst case, ocean speed estimation accuracy was degraded by 0.94% though.
Estimates of precipitation at a high time and space resolution are required for many important applications. In this paper, a new global precipitation map with high spatial (0.1 degree) and temporal (1 hour) resolution using Kalman filter technique is presented and evaluated. Infrared radiometer data, which are available globally nearly everywhere and nearly all the time from geostationary orbit, are used with the several microwave radiometers aboard the LEO satellites. IR data is used as a means to move the precipitation estimates from microwave observation during periods when microwave data are not available at a given location. Moving vector is produced by computing correlations on successive images of IR data. When precipitation is moved, the Kalman filter is applied for improving the moving technique in this research. The new approach showed a better score than the technique without Kalman filter. The correlation coefficient was 0.1 better than without the Kalman filter about 6 hours after the last microwave overpasses, and the RMS error was improved about 0.1 mm/h with the Kalman filter technique. This approach is unique in that 1) the precipitation estimates from the microwave radiometer is mainly used, 2) the IR temperature in every hour is also used for the precipitation estimates based on the Kalman filter theory.