Land surface phenology reflects the responses of a terrestrial ecosystem to climate change. In this study, we used monthly PV (photosynthetically active vegetation) derived from MODIS (the Moderate Resolution Imaging Spectroradiometer) reflectance data between 2001 and 2010 to delineate the land surface phenology. The phenological metrics including onset, offset, and the length of the growing season were obtained from the phenological analysis toolbox TIMESAT. Results indicated that the interannual variability of onset time was significantly correlated to offset time (r = 0.67, p = 0.035) and length of the growing season (r = 0.77, p = 0.009). In addition, the shortage of spring rainfall was influenced by ENSO events and resulted in significant delayed on vegetation onset time and cause dramatic impacts on the ecosystem. TIMESAT is a robust phenological analysis tool, which can facilitate our understanding of the vegetation responses and development associated with climate change and environmental disturbances.

Delineating rice cropping activities is important for crop management and crop production estimation. This study used 8-day the time-series MODIS NDVI data in 2010 to delineate rice-cropping activities in the Lower Mekong Basin (LMB). The data were processed using the wavelet transform and artificial neural networks (ANN). The classification results assessed using the ground reference data indicated the overall accuracy and Kappa coefficient of 84.9% and 0.81, respectively. Comparisons between MODIS-derived rice area and rice area statistics at the provincial level for each country reaffirmed a close agreement between the two datasets (R2 >0.9). The results achieved from the methods used in this study confirmed the validity of mapping approach for delineating rice agriculture activities in the study region. The results could provide quantitative information on rice cropping areas and farming activities. This classification approach is thus proposed for monitoring rice cropping activities in the study region and other regions in the world.

In this study, Landsat-5 satellite images in 1995, 2009 and 2011 were used to calculated the land surface temperature and explore the intensity of the heat island effect. In the results, the intensity is higher in the center of Taichung city and the northwest area. By overlaying with the population map and the industrial map, we found that these areas are the heavy concentration of people and factories. The overdevelopment enhances the urban heat island effect. Considering the different seasons, currents and the wind direction, In the summer, the change of currents and the wind direction make the heat island effect spread to multiple areas. In addition, we explored the heat island intensity in industrial areas. We found that there is no significant correlation between the type of industry and the heat island intensity. The heat island intensity is affected by the wind direction mainly and the higher temperature zone concentrates in the area and leeward.

To collect complete and correct island information, this study mapped South China Sea islands and monitored the natural environment and human developments of important islands. Due to their half-meter resolution capability and 5m accuracy without ground control, high resolution satellite images are suitable for sparse island mapping. The suitability and flexibility of satellite bring certain advantages. Satellites can freely observe the world and surpass ground surveying or aerial photogrammetry for sparse islands in South China Sea. It could be the best option in this study to map islands that are far away from Taiwan. This paper reports a mapping project for islands in South China Sea. Major works include: (1) rectification of satellite images, (2) vectorization, (3) base map generation, (4)3D surface construction, and (5) change monitoring. First of all, rational function model is selected in the generation of orthoimages. Then the vector maps are digitized in the orthoimages. Base maps are generated for analogue output according to the standard format. The next step is to generate digital elevation models by means of image matching and manual editing. Finally, change monitoring is done by the comparisons of historical images. The results provide valuable geoinformation for South China Sea.

Geocenter motion, defined as the motion of center of mass (CM) of the Earth relative to the center of figure (CF), reflects the mass redistribution in Earth system, so accurately estimated geocenter motion has a significant contribution for understanding the mass variation of the Earth. In the study, geocenter motions are determined accurately using the combination of GRACE gravity solutions with altimeter observations or ocean models. The estimates derived from the combination of GRACE and ocean models agree well with previous researches. However, since the altimeter measurements refer to International Terrestrial Reference Frame (ITRF), which is different from CF, and data and models also contain errors, estimated geocenter motions from the combination of GRACE and altimeter data show the discrepancies compared to previous studies, especially in phases and trends. The study also analyzes the impacts of leakage effects and the degree one coefficients of Post-Glacial Rebound (PGR) on determining geocenter motions. The result shows that leakage effects have to be considered in computation. When non-zero degree one coefficient of PGR is applied, only about 80% of the change of assumed degree one PGR reflects on the trend of the estimated geocenter motion.