SUPPORTING GEOTECHNICAL DECISIONS USING RAPID 4D LIDAR CHANGE DETECTION

Managing geotechnical hazards requires geo-professionals to understand how the earth may behave into the future. Understanding morphological change involves a deep understanding of geology, geological processes, climate change, and knowledge of physical changes that have happened in the past or may occur in the future. One of the most capable techniques for mapping changing terrain across spatially extensive regions is lidar change detection using airborne lidar scanning (ALS) data.

Change detection is commonly performed using DEM differencing (DoD), but this introduces major simplifications into the data, potentially limiting one’s ability to detect geomorphic change. Four-dimensional point-based change detection is preferable for its higher precision and lower detection limits, but the computational requirements are much higher compared to DoD. To overcome this limitation, we present a novel GPU-based 4D change detection workflow, including automated chaining of processing steps and dynamic parameter selection. From data acquisition to product delivery, our workflow results in an estimated 200-times increase in efficiency. This enables a wide variety of new applications, including rapid response to disaster events and proactive geohazard screening over thousands of kilometers of infrastructure. Based on these results, we suggest that the standard of practice in applied earth science change detection should move beyond DEM-based methods.