How is tree cover change calculated?
Tree cover change is calculated using data on woody cover from the National Forest and Sparse Woody dataset. The data are an annual time series in which 25 m grid cells are classified as either non-woody, sparse woody (sub-forest woody cover where crown cover is between 5-19%) or forest (woody vegetation ≥2 m tall with crown cover >20%). The phrase “tree cover” is used here to refer to areas with sparse woody or forest cover.
The Explorer calculates the percentages of forest, sparse woody and non-woody pixels in the carbon estimation areas (CEAs) (or project area where CEA data have not been released) for each year from the year the project was registered. Simple linear models are fit separately to the time-series of the forest and sparse woody percentages, with the percentage of each cover class as the response variable and year as the only independent variable. The slope of those relationships are used to classify cover trends inside the CEAs (or project area), provided there are at least three data points in the time series. Projects without three data points receive a “Not enough data” badge.
Projects are given an “Increased” trend badge if the slope of either of the fits for forest or sparse cover is greater than0.25% per year, provided the slope for the fit to the other woody cover class(i.e. sparse if the forest fit has a positive slope >0.25%) is greater than-0.05% per year.
Projects that do not qualify for an “Increased” trend badge receive a “Negligible” trend badge if the sum of the slopes from the linear models fit to the sparse and forest cover are greater than -0.25% per year. All other projects receive a ‘Decreased” trend badge.
The Timeseries tab provides graphs of the data for the three cover classes inside the CEAs (or project area), including the period from 1988 to project registration.