Forest cover, age and densification rate for Southern China at 30 m resolution

The dataset includes Landsat based forest cover maps from 1986-2020 that were used to calculate forest age and densification rate.

Forest probability as the output from a Random Forest model was used to reflect forest cover, and shows how likely an area resembles a dense forest. We set a threshold of 50% probability to define an area as dense forest, and the number of years from the year the threshold is crossed until 2018 as the forest age. Note that the area must remain above 50% until 2018 to qualify as forest. The years before the threshold is crossed are used to calculate the densification rate. It is defined as the mean probability change per year in the period where an area is between a probability of 20% and 50%. If the area falls below 20% the count is reset.