5:15 PM - 7:15 PM
[U03-P03] Monitoring Land Use Changes in Slope Areas—A Case Study of Taoyuan City, Taiwan
Keywords:Land Use Changes in Slope Areas, Satellite Monitoring, Unmanned Aerial Vehicle (UAV) image
In recent decades, Taiwan has undergone a significant economic transformation, shifting from an agriculture-based to an industrial-based society. This transformation has led to significant changes in land use and has accelerated the development of slope areas. To address the issue of illegal land use in slope areas, this study explores the application of satellite imagery in land use change monitoring, using the satellite monitoring project implemented since 2016 by the Department of Water Resources, Taoyuan City Government, Taiwan, as a case study. This project focuses on two main areas: (1) Land Use Changes Detection: Utilizing SPOT6/7 satellite images with a spatial resolution of 1.5 meters for extensive, multi-temporal land use change monitoring, complemented by on-site investigations to verify the legality of detected changes. Additionally, both a hotspot analysis based on illegal cases and an AI-based Random Forest model were developed to identify violation hotspots in slope areas. (2) Multi-Scale Monitoring: In specific areas or identified violation zones, unmanned aerial vehicle (UAV) is deployed to capture high-resolution images and generate 3D models. This enables in-depth analysis of illegal activities and the monitoring of corrective actions for violations.
In land use changes detection, between 2016 and 2024, SPOT6/7 satellite monitoring detected 4,513 changes, of which 1,332 were suspected violations. Using AI-based Random Forest modeling, which integrates illegal cases and spatial factor data, the accuracy of the predictive model improved from 73% in 2016 to 84% in 2024. The analysis indicates that violation hotspots are primarily concentrated in northern Taoyuan City, near the Taipei metropolitan area. This region is predominantly hilly and sloped. In recent years, increasing land development activities have made it a key focus area for monitoring.
Regarding multi-scale monitoring, from 2016 to 2024, high-resolution UAV images and 3D models were generated for 32 areas. Among them, corrective actions have been successfully completed in 13 areas, with vegetation showing good recovery, demonstrating that the remediation measures have achieved the expected results. However, in 19 areas, vegetation has yet to recover, indicating that corrective efforts remain incomplete and require ongoing monitoring. All monitoring results have been uploaded to an online imaging platform, enabling users to query, view, compare images from different periods, and conduct 3D Visualization.
Despite limited funding, the current satellite monitoring project effectively integrates SPOT6/7 satellite images with UAV images, balancing image resolution, monitoring coverage, and costs to optimize resource and time utilization. This study aims to secure additional funding in the future to adopt sub-meter satellite images and increase the frequency of UAV surveys. These advancements will enhance spatial resolution and monitoring accuracy, ensuring a more comprehensive understanding of land use changes in slope areas.
In land use changes detection, between 2016 and 2024, SPOT6/7 satellite monitoring detected 4,513 changes, of which 1,332 were suspected violations. Using AI-based Random Forest modeling, which integrates illegal cases and spatial factor data, the accuracy of the predictive model improved from 73% in 2016 to 84% in 2024. The analysis indicates that violation hotspots are primarily concentrated in northern Taoyuan City, near the Taipei metropolitan area. This region is predominantly hilly and sloped. In recent years, increasing land development activities have made it a key focus area for monitoring.
Regarding multi-scale monitoring, from 2016 to 2024, high-resolution UAV images and 3D models were generated for 32 areas. Among them, corrective actions have been successfully completed in 13 areas, with vegetation showing good recovery, demonstrating that the remediation measures have achieved the expected results. However, in 19 areas, vegetation has yet to recover, indicating that corrective efforts remain incomplete and require ongoing monitoring. All monitoring results have been uploaded to an online imaging platform, enabling users to query, view, compare images from different periods, and conduct 3D Visualization.
Despite limited funding, the current satellite monitoring project effectively integrates SPOT6/7 satellite images with UAV images, balancing image resolution, monitoring coverage, and costs to optimize resource and time utilization. This study aims to secure additional funding in the future to adopt sub-meter satellite images and increase the frequency of UAV surveys. These advancements will enhance spatial resolution and monitoring accuracy, ensuring a more comprehensive understanding of land use changes in slope areas.