Critical aspects of remote sensing of farmland abandonment

Abstract

Remote sensing technology became a powerful tool to monitor various land-change processes, including farmland abandonment, due to the unprecedented ability of remote sensors to capture land-change processes across space and time. Despite the overall progress of satellite, aerial and ground remote sensing of farmland abandonment, data-determined and user-determined gaps do exist. Here we reveal the progress of remote sensing of farmland abandonment, with raising two big questions: 1) where are we in farmland abandonment studies? 2) which critical research gaps do exist? With expert knowledge and a systematic literature review, we elaborate on the progress of farmland abandonment studies, major sensors to be applied, how many existing definitions and produced abandonment maps match the expectations of the users, how diverse mapped farmland abandonment is, among other questions. The systematic review of literature indexed by Clarivate Web of Science suggests that the first very study goes back to 1996 on secondary forest regrowth on abandoned farmlands in tropical regions of Latin America. In general, the literature on farmland abandonment studies increased over time but still is disproportional compared to agricultural expansion and often, the simplified trajectories of farmland abandonment are mapped (e.g., from cropland to abandoned). Despite of changing the paradigm on free accessibility of Landsat imagery after 2008, and the advent of Copernicus programs, multisource remote sensing is not that common, as it would presumably allow for nuancing different stages of natural succession on abandoned farmlands or evolving land uses. We also noticed very few studies follow user-oriented definitions of abandonment rather they develop their own. Also, most studies are at local or regional scales and do not match each other with definition, extent and monitoring period. Also, very few studies employ ground reference data, and rather rely on high-resolution bud subjective image snapshots via Google Earth and Bing. Therefore, we point out several critical directions to fill existing research gaps, including those listed above. Presentation contributes to the activity of Global Land Program’s “Agricultural Land Abandonment as a Global Land-Use Change Phenomenon” https://glp.earth/how-we-work/working-groups/agricultural-land-abandonment-global-land-use-change-phenomenon and https://www.land-abandonment.org/ . The presentation summarizes the research paper in review.