Data and code for: High-Resolution (10-m) National Scale Forest Canopy Height Mapping Using GEDI with Optical and SAR Data

This repository provides all data and codes used to generate and validate the first wall-to-wall high-resolution (10 m) tree canopy height (TCH) map of Bangladesh for 2024. The dataset includes spatial inputs, observation plots used for validation, processed outputs, and reproducible modeling scripts developed within Google Earth Engine (GEE). The spatial data (52 grid tiles shapefile: BGD_Grids_52) are used for national-scale modeling and tiling framework, and the tree/non-tree mask (2024) raster (TreeNonTree2024BD) for final TCH filtering and the final post-processed national TCH raster (TCH_BD2024_TreeOnly5). The validation datasets include Bangladesh National Forest Inventory (NFI) plots (Bangladesh_NFI_Plots_Data and Treeht_BFI15), SPARRSO mangrove field observation plots (SPARRSO_Observation_Plots_52), and Uddin et al.'s (2023) mangrove plot dataset (Sundarbans_Uddin_et_al_data_45). The modelling code (Google Earth Engine, JavaScript) includes a tile-based Random Forest modelling script (Code_Tile_Model.txt), a stratum-based Random Forest modelling script (Code_Stratum_Model.txt), a single national Random Forest model script (spatially balanced sampling) (Code_National_Model.txt), and an additional script for visualization of final TCH outputs (Sample_visualization_BDTCH.txt). These datasets and scripts enable full reproduction of the modelling workflow, from sample extraction and training to validation and final 10m TCH map generation. The repository supports transparent national‑scale forest monitoring, facilitates MRV (Measurement, Reporting, and Verification) applications, and can be adapted for canopy height mapping in other regions.