ICESat-2, a laser altimetry satellite launched by NASA in 2018, was designed to provide global elevation data (ATL08). Together with the existing global DEMs, it can be considered as an alternative for users in Taiwan. To understand the performance of this product, the accuracy of the elevation derived from ICESat-2 ATL08 was firstly evaluated in this study. Then the factors influencing elevation accuracy, including land cover type, slope and altitude, were analyzed based on different spatial units. It was indicated that the elevation error increases as the land cover changes from no vegetation to full vegetation coverage. Overall, elevation error increases with slope and elevation. The results are expected to be helpful for the applications of ICESat-2 elevation data in Taiwan as a reference.

Recent studies have revealed the benefits of green space exposure for schizophrenia. However, vegetation index is mostly used as indices of green space exposure, but in fact there are still many green space-related factors that may be related to mental disorder, such as the size, shape, and proximity of green spaces (called green space structures), but few studies have investigated it, and there is no research to further estimate the lifetime healthcare costs that may be saved after different indices of green space structure are exposed. This research uses the health insurance database, NASA MODIS satellite images, the second national land survey database, geographic information system (ArcGIS) and FRAGSTATS 4 for research. The results of this study found that the greener the exposed green space is, the larger the area, the more complex the shape, and the closer the green spaces are to each other, which may reduce the risk of schizophrenia and may save 0.10 to 0.52 million US dollars in lifetime healthcare costs.

At present, Foehn winds monitoring methods are mainly based on surface weather stations. The weather stations identified as Foehn winds mainly with two indicators: one is temperature, and the other is relative humidity. This study tried to use remote sensing technology and the spatial and temporal adaptive reflectance fusion model (STARFM), which fused 30-meter spatial resolution of Landsat-8 and 10 minutes temporal resolution of Himawari-8 with a high-temporal image fusion method. The STARFM was supposed to improve the problems which is hard to acquire the remote sensing data on the day and the time of the Foehn winds occurrence. The analysis results were expected to achieve the intensive monitoring effects. In terms of temperature, land surface temperature (LST) is generated by using fused thermal infrared (TIR). For humidity, it is difficult to directly observe relative humidity with remote sensing data at this stage, so this study used the related indicators, vegetation indices, which have been widely used for land surface drought monitoring. The results show that for the short-period Foehn phenomenon, the image fusion method can be applied to monitor, but cloud cover is still a difficult factor to overcome. Both the change trend of LST and vegetation indices correspond with the temporal change caused by Foehn winds, and the change of Dawu river basin is especially more obvious than other areas.

Freak wave is a wave with large wave height generated suddenly. It mostly occurs in the sea, however it also occurs frequently in the coast. Many people drop into sea due to coastal freak wave attacked in Taiwan. In the past, most of the studies on coastal freak waves were carried out with numerical simulations because of lack of field data. This study aims to develop a measurement and analysis method for coastal freak wave. It is based on the image process technique. The proposed image process method derives the shape of the splash trigger by coastal freak water. The flow rate and throw speed of the splash were estimated. The parameter Flood Hazard (FH) was used to identify the coastal freak wave. The verification results show that the accuracy of the identification of the coastal freak wave is more than 80%. The result is also agreed with actual accidents reported in media. In addition, the analysis found that the occurrence of coastal freak waves is not only correlated with incident waves. The wave interaction with structure plays an important role for the generation of coastal freak wave. The measurement method for coastal freak waves developed in this study is helpful for establishment of warning system for coastal freak waves.