Mapping Marsh Vegetation and Elevation
PIs: Christine Hladik (Department of Geology and Geography, Georgia Southern University, Stateboro, GA) former affiliation: University of Georgia
Georgia Coastal Management Program, Coastal Incentive Grant
(1) Evaluate the accuracy of a digital elevation model (DEM) derived from light detection and ranging (LIDAR) with real time kinematic (RTK) GPS. (2) Develop species-specific correction factors for 10 marsh cover classes. (3) Use hyperspectral imagery (HSI) to classify salt marsh vegetation and apply correction factors to DEM.
- The unmodified DEM over-predicted elevations and resulted in mean errors of 0.03 to 0.28 m.
- The modified DEM resulted in overall reduction of mean error and slightly under-predicted elevations.
- None of the DEM elevations were significantly different from RTK.
- The author concluded that DEM over-predicts ground elevations. She found that error increases with plant height and is species-specific. These results are consistent with previous studies that reported offsets ranging from 0.07 to 0.13 m for S. alterniflora. Application of correction factors reduces DEM error and greatly improves accuracy.