A novel approach to mapping land conversion using Google Earth with an application to East Africa

作者：Andrew P. Jacobson, Jasjeet Dhanota, Jessie Godfrey, Hannah Jacobson, Zoë T. Rossman, Andrew Stanish, Hannah Walker, Jason Riggio · 发表于：Environmental Modelling & Software · 年份：2015 · DOI：10.1016/j.envsoft.2015.06.011 · 被引用次数：97 · 研究领域：Land Use and Ecosystem Services、Remote Sensing in Agriculture、Species Distribution and Climate Change

Effective conservation planning relies on the accurate identification of anthropogenic land cover. However, accessing localized information can be difficult or impossible in developing countries. Additionally, global medium-resolution land use land cover datasets may be insufficient for conservation planning purposes at the scale of a country or smaller. We thus introduce a new tool, GE Grids, to bridge this gap. This tool creates an interactive user-specified binary grid laid over Google Earth's high-resolution imagery. Using GE Grids, we manually identified anthropogenic land conversion across East Africa and compared this against available land cover datasets. Nearly 30% of East Africa is converted to anthropogenic land cover. The two highest-resolution comparative datasets have the greatest agreement with our own at the regional extent, despite having as low as 44% agreement at the country level. We achieved 83% consistency among users. GE Grids is intended to complement existing remote sensing datasets at local scales.