Cropland capture: A gaming approach to improve global land cover

Abstract

Accurate and reliable information on global cropland extent is needed for a number of applications, e g. to estimate potential yield losses in the wake of a drought or for assessing future scenarios of climate change on crop production. However, current global land cover and cropland products are not accurate enough for many of these applications. One way forward is to increase the amount of data that are used to create these maps as well as for validation purposes. One method for doing this is to involve citizens in the classification of satellite imagery as undertaken using the Geo-Wiki tool. This paper outlines Cropland Capture, which is simplified game version of Geo-Wiki in which players classify satellite imagery based on whether they can see evidence of cropland or not. On overview of the game is provided along with some initial results from the first 3 months of game play. The paper concludes with a discussion of the future steps in this research.