Satellite Imagery
The maps, photographs, and remote sensing images in this atlas were all created with the goal of assessing land cover and land conditions, and measuring change over time.
Maps show a broad view of a country or region, and ground photographs can document landscapes for a particular location, but remote sensing images are the key tools to detect and record surface conditions and understand changes happening on the landscape, both natural and human-caused. Remote sensing images are an objective, cost-effective way to measure and analyze long-term change, including the change in land cover from 1975 to 2013 that is at the heart of this book.
Some of the images are taken from Google Earth and credited as such, but the majority come from one of three sources: Landsat, Corona, or the Moderate Resolution Imaging Spectroradiometer (MODIS). Each has its own characteristics and advantages.
The first Landsat satellite was launched in 1972, and the program has been in continuous operation since then. Landsat was designed specifically to study and map land resources. Landsat 8 now orbits the Earth at an altitude of 705 km and records data as 30-m pixels in images approximately 170 km by 185 km. Landsat images revisit the entire Earth every 16 days.
Corona was a national reconnaissance mission, flown on satellites from 1960 to 1972. Corona recorded photographs on high quality film stock, which was jettisoned and recovered in the atmosphere by airplane. The high quality fi meant that Corona photos recorded fine details, but coverage was limited to areas of interest to U.S. military programs during the Cold War. Nevertheless, Corona photography of West Africa covers virtually all of West Africa, dating back to as early as 1962. Corona film was declassified in the interest of science in 1995. Project staff at the U.S. Geological Survey (USGS) EROS Center coordinated the scanning and georegistration needed to convert Corona film photography into digital image data.
The MODIS instrument is mounted on the Terra and Aqua satellites, launched in 1999 and 2002, respectively, that orbit at the same altitude as Landsat. However, MODIS images cover a swath 2,330 km wide at a relatively coarse resolution of 250 m. As a result, MODIS images are less detailed than Landsat, but instead of every 16 days, MODIS has the advantage of covering the entire Earth every one or two days. MODIS provides the data for calculating a widely used index of vegetation condition, the normalized difference vegetation index, or NDVI.
The Landsat program has served longer than any other Earth-observing satellite system. For that reason, Landsat provides the bedrock dataset for land cover mapping and land cover change analysis for most of the maps in this book. Landsat is detailed enough at 30-m resolution to map and measure many types of landscape changes, for example the growth of agriculture and of cities, as well as the fragmentation of forests and savannas. Landsat’s 16-day repeat cycle is frequent enough to make it possible to overcome the frequent heavy cloud cover in some parts of West Africa. The consistency of Landsat imagery makes it possible to make objective observations of land cover change from 1972 to the present.
Corona serves the important function in several areas of pushing the observation window back another 10 years before Landsat, with satellite photography from the early 1960s that complements Landsat imagery. MODIS data at 250-m resolution serve as a base for several national and regional maps in this atlas and for assessing vegetation condition. The differences in footprint and visual characteristics among the three systems can be seen in the comparison above.
Mapping Land Use and Land Cover
