Objective: Predict annual forests and therefore temporal trends of forest loss. Attribution and filtering of forest losses due to natural causes.
Method:
Random Forest Regression for Woody Coverage Prediction;
Scientifically Informed Thresholding (Coverage -> Crown Covers -> Forests);
False Positives Removal with Ancillary Bushfire Datasets
Example results:
Metrics:
