|
Mangroves are coastal vegetation and considered as one of the blue carbon ecosystems. Canopy height is a critical parameter in understanding these ecosystems and related to biomass information. Remote sensing and machine learning techniques are increasingly utilized for large-scale mangrove canopy height mapping. Sentinel-2 is an optical satellite with various bands, including red-edge (RE) bands that are related to vegetation biophysical information. This study investigates the potential of red-edge spectral indices for mangrove canopy height mapping using Random Forest Regression (RFR). The study area is located around Charlotte Harbor Preserve State Park, Florida, USA in 2020. Sentinel-2 data was used to produce several red-edge spectral indices including NDVI, NDVIRE1, NDVIRE2, NDVIRE3, CIRE1, CIRE2, CIRE3, and CIVI which served as inputs for the RFR model. Spaceborne GEDI LiDAR rh98 canopy height data was used here to produce the target data. The dataset was divided into 80% training and 20% testing subsets. Our results indicate that CIVI utilizing red-edge 1 and 2 is the most important feature of RFR for mangrove canopy height estimation. The mean absolute error (MAE) and the root mean squared error (RMSE) based on the testing dataset were 1.662 m and 2.291 m, respectively. To further validate the results, we introduced an independent testing dataset using a canopy height model from the airborne LiDAR data located near the study area. We used the trained RFR model to predict the mangrove canopy height in the testing area and found the MAE and RMSE from the independent testing dataset were 2.511 m and 2.812 m, respectively. Based on the results, most of the red-edge spectral indices have a better performance than the spectral index without red-edge bands, and the use of several red-edge spectral indices provides promising results for mangrove canopy height mapping.
|
