Exposing to Nitrogen Dioxide (NO2) may cause adverse health effects. Previous air pollution estimating models still face overfitting or underfitting problems. Thus, using estimation model with high spatial and temporal resolution to assess NO2 exposure is important. This study utilized Land-Use Regression (LUR) coupled with Extreme Gradient Boosting (XGBoost) algorithm to feature NO2 concentration distribution in Kaohsiung metropolis. The results showed that R2 value for LUR integrated XGBoost model was 0.82, RMSE was 4.53 ppb, which had highly explanatory ability. Besides, 10-fold cross validation R2 for the proposed model was 0.82, which showed that the model did not encounter overfitting issue. Finally, this study used the model to depict estimation maps for NO2 concentration variation in Kaohsiung. The results showed that higher polluted regions were clustered in south Kaohsiung where industries were well developed and population was densely distributed.

The indigenous peoples have been living in Taiwan for thousands of years. Comparing to the western-style science education, they have completely different pedagogy to establish and to pass their own environmental knowledge, living styles, and cultures. However, the young generation of the indigenous people is losing their traditional point of view to their environment under current western-style science education system. Especially for the indigenous peoples who moved to cities for living, the opportunities for their young generation to learn how to interact with natural environmental resources are significantly decreased. Based on the above, we proposed a concept called “Virtual Indigenous Tribe” in this study. We integrate traditional ecological knowledge into Immersive Virtual Reality to let urban indigenous youths learn the knowledge in a more intuitive way. We collected traditional ecological knowledge through the interviews with elders, integrating it into panoramic videos and images taken by 360° panoramic cameras and the Unmanned Aerial Vehicle (UAV) to build Virtual Indigenous Tribe. Finally, we use Unity to develop immersive virtual tribe APP. We cooperate with urban indigenous class of Shulin Senior High School to evaluate the effectiveness of Virtual Indigenous Tribe teaching. The results show that Virtual Indigenous Tribe teaching can significantly improve more students' cognition of traditional ecological knowledge than general teaching. In this study, we hope that “Virtual Indigenous Tribe” can be used as an auxiliary teaching tool to enable urban indigenous students to break through the limitations of time and space to learn traditional ecological knowledge in a more authentic way, and increase students’ curiosity about their own tribal history, culture and knowledge.

This study proposes a novel learning-based point cloud structurization method especially for building components such as plates, beams, columns, and walls. The proposed model net learns from the point clouds generated by existing building information modeling (BIM) components to predict the geometric model of a newly given point cloud consequently. It is worthy to note that the BIM-to-Point cloud approach overcomes the difficulty of 3D training data acquisition that is usually confronted in most deep learning applications. The average precision (AP) of each category achieved a precision of above 50% in the classification task, in which the wall category yielded a precision of 77%. On the other hand, the predicted object corner positions reported an accuracy better than 25cm, in which the beam category revealed an accuracy of up to 10cm. Moreover, the quality of the geometric model in vector-linking prediction reached a precision of 70% suggesting that the proposed learning-based method can indeed reconstruct the geometric model of a given point cloud automatically.

Water quality monitoring is nowadays an essential environment issue. Compared to the traditional survey, using remote sensing data to estimate the water quality in the reservoir is more time-saving, efficient in a large area. This study utilized the satellite images of FormoSat-5 (FS5), Landsat 8 (L8), and geographic information system (GIS) to monitor the water quality of Shimen and the Baoshan Reservoir. The predicted regression model was built based on the spectral data of the collected satellite images and ground-referenced data of the water quality of the Shimen Reservoir and Baoshan Reservoir. The R2 of chlorophyll-a (ChIa), secchi disk depth (SDD), chemical oxygen demand (COD), and total phosphorus (TP) in Shimen and Baoshan Reservoir are between 0.61 and 0.75. The mapping results can not only present the spatial distribution of water quality in the reservoir area, but also indicate that water quality has high correlation with the land use in the reservoir coastal area. In this study, the procedure of combining remote sensing data and the regression model to estimate the spatial distribution of water quality of the reservoir can be promoted to monitor the water quality of other reservoirs in Taiwan in the future.