Climate - in terms of temperature, precipitation and continentality - is a primary determinant in the distribution of African vegetation / ecosystems. Salvador Rivas-Martinez and Salvador Rivas-Saenz (2004) developed a global bioclimatic classification system that quantifies key bioclimatic indices reflective of vegetation distributions. These indices can be used to model thermotypes (i.e. hot-cold gradients), and ombrotypes (i.e. wet-dry gradients). Their model was translated into GIS spatial algorithms to produce bioclimate data for the US ecological systems mapping project (Sayre and others, 2009). These spatial models were used (with minor adaptations) with Worldclim climatological data (Hijmans et. al. 2005) to model/map African ombrotypes and thermotypes. These two datasets were then combined to produce an isobioclimate map with a total of 157 composite classes.

The African isobioclimate data was developed as a primary input dataset for an African Ecological Footprint mapping project undertaken by the U.S. Geological Survey and The Nature Conservancy. The project used a biophysical stratification approach - combining isobioclimate, surficial lithology, land surface forms, landcover, topographic moisture potential, and biogeographic ecological divisions - to generate ecological footprints. The composition and distribution of these unique footprints of the physical and biological landscape was then reviewed by regional vegetation and landscape ecology experts and attributed (labeled) to an intermediate scale African ecosystem class.

The land surface forms were identified using the method developed by the Missouri Resource Assessment Partnership (MoRAP). The MoRAP method is an automated land surface form classification based on Hammond’s (1964a, 1964b) classification. MoRAP made modifications to Hammond’s classification, which allowed finer-resolution elevation data to be used as input data and analysis to be made using 1 km2 moving window (True, 2002; True and others, 2000). While Hammond’s methodology was based on three variables, slope, local relief, and profile type, MoRAP’s methodology uses only slope and local relief (True, 2002). Slope is classified as gently sloping or not gently sloping using a threshold value of 8%. Local relief is classified into five classes (0-15m, 16-30m, 31-90m, 91-150m, and >150m). Slope classes and relief classes were subsequently combined to produce eight land surface form classes (flat plains, smooth plains, irregular plains, escarpments, low hills, hills, breaks/foothills, and low mountains). Sayre and others (2009) further refined the MoRAP methodology to identify a new land surface form class, “high mountains/deep canyons”, by using an additional local relief class (>400 m). The Africa implementation used a 90-meter elevation dataset which was created by void-filling and re-sampling the 30-meter SRTM elevation data provided by the National Geospatial- Intelligence Agency.

In the preliminary output, artifacts were identified over flat desert areas affecting the classification between the two lowest relief classes, “flat plains” and “smooth plains.” Since this problem was especially pronounced in areas where the SRTM data originally had data-voids, the problem could have been caused by anomalies or artifacts in the input elevation data, which resulted from void-filling processes. Instead of further investigating causes of the problem, the two land surface form classes were combined. In addition, the “low hills” class which had a very low occurrence was combined with the “hills” class. As a result, seven land surface form classes were identified in the final dataset.

The Topographic Position dataset identifies two classes: uplands, lowlands/depressions. This is a derivative product of the Compound Topographic Index (CTI) data produced from the 90m Shuttle Radar Topography Mission (SRTM) elevation data. The CTI is a topographically derived measure of slope for a raster cell and the contributing area from "upstream" raster cells, and thus expresses potential for water flow to a point.

The Topographic Position classes were identified by grouping CTI values into two classes using a threshold which represented the interface between uplands and lowlands. The determination of this threshold required an additional dataset, the SRTM River-Surface Water Bodies dataset, which mapped inland water body boundaries. These water body boundaries were overlain on the CTI data. CTI values which occurred over the boundaries were extracted, and the frequency distribution of these values was calculated. The mean value in this distribution was found to sufficiently identify the uplands-lowlands interface and was used as the threshold.

The African surficial lithology dataset is a map of parent materials - a mix of bedrock geology and unconsolidated surficial materials classes. The goal was to produce a map that reflected the key geological parent materials which act as primary determinants in the distribution of African vegetation /ecosystems. It is a compilation and reclassification of twelve digital geology, soil and lithology databases. Nineteen surficial lithology classes were delineated in Africa based on geology, soil and landform. Whenever available, multiple sources of ancillary digital data, hard copy maps and literature were reviewed to assist in the reclassification of the source data to the African surficial lithology classification. Of particular note, due to the varying spatial and classification resolutions of the geologic source data, the African surficial lithology map varies in spatial complexity and classification detail across Africa.

The African surficial lithology data was developed as a primary input dataset for an African Ecological Footprint mapping project undertaken by the U.S. Geological Survey and The Nature Conservancy. The project used a biophysical stratification approach - combining isobioclimate, surficial lithology, land surface forms, landcover, topographic moisture potential, and biogeographic ecological divisions - to generate ecological footprints. The composition and distribution of these unique footprints of the physical and biological landscape was then reviewed by regional vegetation and landscape ecology experts and attributed (labeled) to an intermediate scale African ecosystem class.

"The input layers described and depicted above for landforms, lithology, bioclimates, thermotypes, ombrotypes, and biogeographic regions were used as predictor variables in the CART analysis. In addition to their utility for predicting ecosystem distributions, these datalayers have considerable potential in and of themselves for a variety of other applications ranging from agricultural planning to biodiversity analysis. With the exception of the biogeographic regions layer, which is a generalization of existing ecoregional datalayers, each input layer represents the finest spatial and classification (thematic) resolution, continent-wide dataset yet available for that theme" (Sayre, 2013).