As noted in the Strategic Plan for the US Climate Change Science Program (CCSP), the systematic measurement of trends in global land cover is a priority for understanding land/climate interactions and maintaining ecosystem goods and services for society. In particular, repeated observations of land cover on time scales of 1-5 years are required to adequately characterize trends such as deforestation, agricultural expansion, urbanization, and ecosystem disturbance.

In the late 1990's, NASA and MDA Federal partnered to create the highly successful Geocover data set. This orthorectified, global dataset of cloud-free Landsat imagery, centered on 1975, 1990, and 2000 epochs, has supported numerous science studies and mapping applications. Geocover data may be obtained at no cost from the USGS EROS, the University of Maryland Global Land Cover Facility, or the Michigan State University Tropical Rain Forest Information Center.

Preferred season for MDGLS imagery, as determined by analysis of the AVHRR NDVI record. Green=northern hemisphere (NH) Summer; Red= NH Spring; Purple= NH Fall; Yellow= NH Winter; Black = no preferred season.

Example of the 2000 Geocover mosiac for the United States (courtesy of MDA Federal).

The MDGLS continues this record by providing a new set of imagery centered on 2005. Cloud-minimized Landsat-7 and Landsat-5 imagery from peak growing season conditions are being acquired from the archives at USGS EROS and international ground stations. The data will then be calibrated to at-sensor radiance and orthorectified to a high-resolution (SRTM) map base. In addition, Landsat-7 ETM+ data will be “gap filled” to remove gaps caused by the failure of the Scan-Line Corrector mechanism since 2003. More details on the product specification can be found in the Documents section.

The MDGLS project is being managed by USGS and NASA. USGS is responsible for identification and acquisition of relevant data (“Phase I”), and processing the data set (“Phase II”). NASA is responsible for “Phase III”, which will involve generation of global land cover products from the MDGLS data sets. The project also solicits scientific direction from a Science Steering Committee composed of government and academic land remote sensing experts.