The IKONOS data were adopted to simulate ROCSAT-2 image. Firstly, a 1*1 m^2 fused image was produced using the IKONOS XS and Panchromatic imagery. Next, the image was geometrically rectified using GCPs, existing DEM and contour maps. Then, the Point Spread Function of NOAA AVHRR 7 was adjusted to resample the IKONOS ortho-image and generate the simulated ROCSAT-2 data with the spatial resolution of 2*2 m^2, which was manually interpreted and the land use/cover GIS information was extracted. The results were compared with that obtained from SPOT image. It shows that fifteen categories under Classes Ⅰ and Class Ⅱ of Land Use and Cover Classification are identified from simulated ROCSAT-2 and SPOT-4 image data. The areas of rice paddy identified from the two image data are in good correlation, but the areas of arid farmland differ by about 30%. Additionally seven categories under Class Ⅲ can by identified from the ROCSAT-2 data only. Most significant is the accuracy of crop identification, which achieves the critical value of 90%. Consequently, ROCSAT-2 will be a good data source not only for GIS data extraction but also for updating the existing paddy and other GIS maps.

Determination of aerial photo orientation is very fundamental issue in photogrammetry. Along with the developments of computer science and join in image-processing technology, many works in digital photogrammetry can be automatically accomplished. As for smaller area, we can update changed terrain features by using the existing data not by using AT on analytical plotter that should be saving much time and human resources. Concerning some partial changes in aerial images, we used Monte Carlo method with area-based matching technique and design a semiautomatic operation procedure which can perform on PC. By determination of aerial image orientation and DTM, we can update the existing orthoimage. And the accuracy of experience result is equal to the accuracy of digital photogrammetry.

The stand height growth of ecosystem can be indicated the vegetation dynamics and site quality, so to discuss this issue is very important for forest management and vegetation conservation. This study using aerial photos of three representative years (1976, 1989 and 1997) as materials, then making the digital surface model (DSM) by using the digital photogrammetry technique. And in order to discuss the impacts of different factors, using the 't-test' to compare the stand height growth of different disturbed area in Nanjenshan Ecological Reserve. The RMS errors of aerotriangulation in this research are as follows: X- coordinate is 1.09m, Y-coordinate is 1.28m, Z- value is 1.45m (1976); X- coordinate is 0.56m, Y- coordinate is 0.89m, Z- value is 1.14m (1989); X- coordinate is 0.27m, Y- coordinate is 0.54m, Z- value is 0.89m (1997). Although the results as above are not in precision, but for large-scale natural resource management, these data are good enough. Otherwise, for discussing the dynamics of stand height growth, using the long interval data (1976 to 1989 and 1976 to 1997) is better than using the short interval data (1989 to 1997). The results of t-test also show that the different factors got different influences in stand height growth. So, when making management strategy, the administration should consider every situation. If the results of this research can be integrated with the dynamics of vegetation physiognomy and the dynamics of canopy density, it will be helpful for the management of Nanjenshan Ecological Reserve.

In the last 21 years, the study area has been influenced by population increasing and the full extent of urbanization. Owing to the aerial photograph has higher spatial resolution than other image data, and in different land use type has diverse image characteristic presenting, it is suitable for land use change detection tools. We used the three period aerial photograph (1976, 1987, 1997) of Tapeng Bay, processing with scanning, ortho-rectification, image assembling, photo interpretation, digitizing, overlay analysis, and assist with GIS facility to monitoring the temporal, spatial land use change detection during the past 21 years. The results show that from 1976 to 1997, the area of fish ponds and urban are increasing 242.05ha and 75.81ha respectively; vegetation decreasing 300.55ha, and beach line had shift condition. It is ought to be pay attention that the Tapeng Bay is a arenaceous coast, in early change there are plenty of mangrove,and because of the dramatic land use change there are fragmentary distribution now.

This paper presents a new supervised classification technique for hyperspectral imagery, which consists of two algorithms, referred to as greedy modular eigenspace (GME) and positive Boolean function (PBF). The GME method is designed to extract features by a simple and efficient GME feature module. The GME makes use of the data correlation matrix to reorder spectral bands from which a group of feature eigenspaces can be generated to reduce dimensionality. It can be implemented as a feature extractor to generate a particular feature eigenspace for each of the material classes present in hyperspectral data. The residual reconstruction errors (RRE) are then calculated by projecting the samples into different individual GME-generated modular eigenspaces. The PBF is further developed for classification. It is a stack filter built by using the binary RRE as classifier parameters for supervised training. It implements the minimum classification error (MCE) as a criterion so as to improve classification performance. It utilizes the positive and negative sample learning ability of the MCE criteria to improve classification accuracy. The performance of the proposed method is evaluated by MODIS/ASTER airborne simulator (MASTER) images for land cover classification during the Pacrim II campaign. Experimental results demonstrate that the GME feature extractor suits the nonlinear PBF-based multi-class classifier well for classification preprocessing. The proposed approach is not only an effective method for land cover classification in earth remote sensing but also dramatically improves the eigen-decomposition computational complexity compared to the conventional principal components analysis (PCA).

The operating frequency of satellite communications is continuously increased to gain wider bandwidth and smaller size of antennas. However, when the operating frequency is higher than 10 GHz, the quality of the earth-satellite communication links may be seriously degraded by rain fading effects. Therefore, how to predict and compensate the rain fading effects accurately is a very important issue for satellite communications. Presently, all the popular rain fade prediction models are assumed to be functions of rain rate. Hence, to predict rain rate accurately is the first step to get an accurate rain fade prediction. This paper is devoted to compare the applicability of all proven rain rate prediction models upon applying the rain rate data measured in Taiwan. Experimental results indicate that Rice-Holmberg model is the most suitable rain rate prediction model in Taiwan area.