GLCF Launches Global Landsat Tree Cover Dataset
The Global Land Cover Facility (GLCF) launched the first 30 meter Landsat based Tree Cover dataset. As part of the Global Forest Cover Change project funded by NASA's MEaSUREs program, these data provide the foundation for monitoring changes in Earth's forests at a pixel resolution of 30 meters - 8 times finer than any previously available tree cover dataset. For more information about the dataset please visit the data and product page for the Landsat Tree Cover dataset.
Based on 30-meter-resolution Landsat images and sampling from the 250-m MODIS VCF Tree Cover and Cropland Probability layers, the GLCF team inverted the traditional, bottom-up method of generating land cover datasets and used these coarser datasets as “parents” with 30-m resolution Landsat images to produce the first of a new generation of high-resolution tree cover datasets. For more information about the methodology used to derive the product, please download the published paper in the International Journal of Digital Earth.
The scientific utility of measurements stems from more than just their scale, however. The data must also be precise and accurate, and this certainty must be known quantitatively. The GLCF used laser imaging to test the accuracy of both the new dataset and its parent. After calibration, the Landsat-based tree cover dataset showed an improved accuracy over its MODIS parent: whereas the MODIS VCF had a post-calibration uncertainty of 14% cover, the Landsat data showed a post-calibration uncertainty of 9%. Thanks to the input of the Cropland Probability Layer, improvements were greatest in agricultural regions—an important consideration not only for monitoring forests, but also mapping and monitoring the planet’s farmland.
Most changes in land cover—deforestation, reforestation, parcel development, etc.—occur in patches smaller than the 250-m pixels resolved by MODIS. This first global, Landsat-based dataset is thus the first step into a new era of high-resolution, global land cover research. The data’s potential is already being realized as the GLCF is now using the data to analyze global patterns of forest cover and change between 2000 and 2005. Further, the GLCF maintains an open-access policy toward data, and requests for the data have already begun pouring in from around the world.