Remote sensing scientists have employed the free and open Landsat archive to produce pathfinding global-scale data sets characterizing both land cover extent and change. Here, we propose a method to consistently characterize annual time-series of global percent tree canopy and bare ground cover, and surface water extent. We will combine indirect post-characterization comparison and direct change detection methods to map land changes. Area estimates of loss and gain for each theme will be made from stratified probability-based samples. The sample-based area estimates will then be used to adjust the map products. Per pixel adjustments using fuzzy logic will enable the matching of total area of mapped change with the sample-based area estimates. In so doing, we will ensure that annual land cover extent maps are consistent with inter-annual land cover change.