Spatial Database Description

The CPC Ecuador Projects have an extremely large database of geospatial data. The database contains a wide variety of vector and raster data sets, including roads, hydrography, and digital elevation models (DEM). It also contains an extensive time-series of satellite imagery, covering thousands of square kilometers. Some of the data sets cover the entire study area, while others cover smaller areas, such as the Intensive Study Areas (ISAs). Four ISAs were created to provide the investigators with a variety of zones within the larger study area in which to perform analyses that would require more time and processing capacity for the complete study area.

Below is a brief description of the contents of the database. For more detailed information regarding any of the data sets that are available for distribution, see the Spatial Metadata page.


A large number of ARC/INFO coverages were created initially by digitizing features from 1:50,000-scale topographic maps. A total of 86 topographic maps, in several different levels of completion, were obtained to cover the majority of the study area. From these maps, we digitized 50 road coverages, 86 hydrography coverages, 16 elevation contour coverages, and 21 point elevation coverages. The difference in the number of each coverage can be attributed to 1) there only being 50 topographic maps with roads; 2) the extreme amount of time needed to digitize contours, resulting in the digitization of contours only for the ISAs; and 3) the digitization of point elevations for the same maps as the contours, plus a few extra.

Another source of data is the Global Positioning System (GPS). GPS receivers were used to collect geodetic control for the rectification of satellite imagery, land use/land cover (LULC) data for the classification of satellite imagery, and house and finca boundary locations for the delineation of finca madres.

The GPS points for the finca front property corners were used in conjunction with satellite imagery and air photos to create a polygon coverage representing the spatial area of each surveyed finca madre.

Other vector data includes political boundaries for the nation of Ecuador and its provinces, the major cities and small communities in the northern Oriente, and smaller-scale (1:250,000) hydrography for the northern Oriente.

Non-Imagery Raster

The majority of the non-imagery raster data falls into one of two categories: topography and land use/land cover (LULC). The topography data includes digital elevation models (DEMs) and some derivative data sets (slope, aspect, and topographic moisture index) for each ISA. The LULC includes a generalized, four-class classification for each ISA for several dates.

Remotely Sensed Imagery

Remotely sensed imagery comprises the richest amount of spatial data in the entire database. The majority of images are from the Landsat Multispectral Scanner (MSS) and Thematic Mapper (TM) satellite-based platforms. Both sensors are passive, meaning that they capture reflected electromagnetic radiation off of the surface of the earth, or off of clouds, if they happen to get in the way. These satellite images provide a regional view of the study area, with each image covering approximately 185km east-west and 200km north-south. The Landsat time-series provides us with snapshots of this area for 30 years, beginning in 1973 and continuing up to the present. The satellite passes over the same location every 16 days, giving users the ability to regularly monitor changes in the LULC. Unfortunately, the climate in this region causes there to be many overcast days, drastically reducing the number of cloud-free images in this 30-year time span from a potential 680 down to a mere 24.

In addition to the Landsat imagery, we have begun acquiring high-resolution IKONOS satellite imagery. IKONOS is a privately-owned passive satellite sensor, operated by Space Imaging in Colorado. IKONOS imagery has a spatial resolution of up to 1 meter (compared to Landsat TM’s 30m resolution), allowing for much more detail to be gleaned from the images, at the cost of a much smaller areal coverage.

Aerial photographs are the third type of remotely sensed imagery in the database. The air photos are from two dates: 1979 and 1990. They have been scanned at 300dpi and can be used to analyze and verify LULC at those two dates.