Global Tree Cover Estimation Algorithm

Global Tree Cover Estimation Algorithm
  • Status Complete
  • Partners Global Good (co-founded by Bill Gates and Intellectual Ventures)

Project Overview

In collaboration with Global Good (co-founded by Bill Gates and Intellectual Ventures), we are developing a tool to evaluate percentage of tree cover from remote sensing imagery to provide the information needed by these implementation and monitoring organizations.

Though intended for application on degraded, reforested, and mixed-use agroforestry land types, the method can be applied in other environments and for other purposes, e.g. quantifying tree coverage as input to tree carbon or soil-water-ecosystem restoration models.

Methodology

The information can be mapped to land cover classifications, and time series imagery can be analyzed for temporal trends. The algorithm is trained using observations from Digital Globe high resolution 1-ha images with tree densities ranging between 0 and 98%, making the method appropriate for regional assessments with a variety of land conditions. The accuracy of the algorithm is assessed against the current dot-grid method, which varies as a function of tree coverage.