Is this project an undergraduate, graduate, or faculty project?
Graduate
individual
What campus are you from?
Daytona Beach
Authors' Class Standing
Joelle Bobinsky, Graduate Student
Lead Presenter's Name
Joelle Bobinsky
Faculty Mentor Name
Dr. Stephen C Medeiros, PhD, PE
Abstract
Lidar-derived digital elevation models (DEMs) are crucial for modeling salt marsh evolution, forecasting storm surge inundation depth and duration, and simulating the coastal dynamics of sea level rise (SLR). Advances in lidar acquisition and data processing techniques over the last decade have led to increased accuracy. However, in densely vegetated coastal salt marshes, lidar-derived DEMs are generally unreliable without adjustment. In this presentation, the researchers investigate the need for local topographic ground truth data to train Random Forest (RF) DEM adjustment models for two similar Northern Gulf of Mexico salt marshes. Two GNSS-RTK field surveys were conducted to acquire ground truth topographic elevations near St. Marks, Florida (n=377) and Pascagoula, Mississippi (n=610). These elevations, along with Sentinel-2A MSI reflectance values, were used to train the RF salt marsh DEM adjustment models under three scenarios: local, non-local, and combined. The RF-local models achieved better MAE values for St. Marks and Pascagoula, respectively. The predictions using non-local data for training were far worse and the predictions using combined ground truth data were marginally worse. The evidence suggests that local ground truth data are necessary for mitigating bias in salt marsh lidar DEMs, although it remains to be seen if substantially increasing the data set size could narrow the accuracy gap.
Did this research project receive funding support from the Office of Undergraduate Research.
No
Locality of Topographic Ground Truth Data for Salt Marsh Lidar DEM Elevation Bias Mitigation Using Random Forest
Lidar-derived digital elevation models (DEMs) are crucial for modeling salt marsh evolution, forecasting storm surge inundation depth and duration, and simulating the coastal dynamics of sea level rise (SLR). Advances in lidar acquisition and data processing techniques over the last decade have led to increased accuracy. However, in densely vegetated coastal salt marshes, lidar-derived DEMs are generally unreliable without adjustment. In this presentation, the researchers investigate the need for local topographic ground truth data to train Random Forest (RF) DEM adjustment models for two similar Northern Gulf of Mexico salt marshes. Two GNSS-RTK field surveys were conducted to acquire ground truth topographic elevations near St. Marks, Florida (n=377) and Pascagoula, Mississippi (n=610). These elevations, along with Sentinel-2A MSI reflectance values, were used to train the RF salt marsh DEM adjustment models under three scenarios: local, non-local, and combined. The RF-local models achieved better MAE values for St. Marks and Pascagoula, respectively. The predictions using non-local data for training were far worse and the predictions using combined ground truth data were marginally worse. The evidence suggests that local ground truth data are necessary for mitigating bias in salt marsh lidar DEMs, although it remains to be seen if substantially increasing the data set size could narrow the accuracy gap.