There is a newer version of the record available.

Published March 27, 2026 | Version v1

Spatially-aware machine learning predictions of tropical forest structure using GEDI metrics and multi-source Earth Observation data

  • 1. Mathematics & Geospatial Science, STEM College, RMIT University
  • 2. Faculty of Geoinformation Science & Earth Observation, University of Twente

Description

The raster layers provided here were gererated from wall-to-wall predictions of GEDI metrics forest structure characteristics (canopy height and cover) of a diverse tropical landscape in Papua New Guinea (PNG) using spatially-aware machine learning modelling, synergising multi-source earth observation (EO) data particularly, Sentinel-1 SAR, Sentinel-2 Optical, AlphaEarth Satellite Embedding, environmental variables and synthetic spatial features. In addition, the CSV files contain the out-of-fold (OOF) prediction data, overall prediction metrics, and the feature importance for each model prediction.

The paper is currently under final review by the GIScience & Remote Sensing Journal, Taylor & Francis. As part of the Journal's Data Sharing Policy (Taylor & Francis open + FAIR data policy), this dataset is being openly shared under CCBY & CC0 Licenses.

Paper Title: Tropical forest structure mapping using spatially-aware machine learning with GEDI metrics and multi-source earth observation data

Paper Highlights

  • Introduction of a mean GEDI canopy height metric (RHm) to minimize inter-percentile sensitivity
  • Development of a Spatially-Aware LightGBM  (SA-LGBM) optimization framework (with nested Bayesian Optimizer)
    • Local Outlier Factor thresholding
    • Eigenvector Spatial Filtering
    • Buffered 5-fold Spatial Block Cross-validation
  • Creation of  a comprehensive forest structure index (FSI) from GEDI multi-metric model predictions
  • First spatially-explicit GEDI-derived forest structure wall-to-wall mapping effort specific to PNG

 

Connect on LinkedIn: Glen R Yali 

Files

FeatureImportance_SpBCV_FHD_Model.csv

Files (1.9 GB)

Name Size
md5:bdfc47beb97a993c8ba6833425eb97fa
4.8 kB Preview Download
md5:be97fd15804808cc28561e306b0a9547
4.8 kB Preview Download
md5:c7b124f359aac7e2c5c3b688b506e957
4.8 kB Preview Download
md5:7196f534af33ae6b8e05b1da658bf8fd
4.8 kB Preview Download
md5:57e1eb9f3bcc203843a28fc6770219c7
371.7 MB Preview Download
md5:49231bd6beed5dc4cb1055864a5a3ddb
371.6 MB Preview Download
md5:847347db494a71df879e51e9216b8f69
371.6 MB Preview Download
md5:8fee79e7085b52f11975e501208e4c28
371.7 MB Preview Download
md5:c0f1009d797ebec547a9f14a80ab7178
371.7 MB Preview Download
md5:2b2c88a1a11ffafac0bb7b9edafe4eae
816.5 kB Preview Download
md5:50d4f60e31c065267ca6ec4b4180bb67
316 Bytes Preview Download
md5:289087480376dab5c23967fcb03b76a7
309 Bytes Preview Download
md5:e77bedc18dd67eb0f28ac6b834dc542d
300 Bytes Preview Download
md5:7ca370500e724ed01ae25c4f7849fcb7
318 Bytes Preview Download
md5:e247c8f38b73cf68a9446d088c71cae9
807.7 kB Preview Download
md5:42b2ca13246babbd5e91d651a6cb5999
788.6 kB Preview Download
md5:2099618b8323a47db7b4fc4bd01b1a48
820.3 kB Preview Download

Additional details

Software

Repository URL
https://github.com/gryali/Geospatial-Machine-Learning
Programming language
Python
Development Status
Active