This page describes the coverage of the global variable datasets used in GLOBE.

LandScan 2007 population density was used as a land mask to select all Global Land Units (GLUs) that contained land. This created a base extent and resolution (30 arcsec) to which all global variable datasets were converted. If the native extent of a global variable was less than that of LandScan 2007 (typically small islands were not included), and it was appropriate and/or possible to extend the variable (e.g. mean temperature can be extrapolated, whereas world protected areas cannot), then the global variable was extrapolated to include as much land mass as possible. A tiered classification was developed that summarizes the final coverage (i.e. extent) and degree of extrapolation applied to the native data.

Any given GLU might contain native values for all global variables, extrapolated variable values, or no data. The smaller the amount of land covered by a GLU, the more likely it is that certain global variables were extrapolated and/or missing. A report of the percent of GLUs that contain data for the given number of global variables (for which global coverage is possible and/or appropriate) is given below.

Global Coverage Report

Nearly 90% of GLUs are missing data for 2 or fewer global variables, and 86% of GLUs have data for all variables intended to have global coverage.

Coverage reports for individual global variables are provided on each variable’s documentation page.