Remote sensing of swidden agriculture in the tropics: A review

• Published in: International journal of applied earth observation and geoinformation Vol. 112; p. 102876
• Main Authors: Jiang, Ningsang, Li, Peng, Feng, Zhiming
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 01.08.2022; Elsevier
• Subjects: biodiversity; Central Africa; climate change; Dry season; ecosystems; evolution; forests; global carbon budget; Landsat; Latin America; shifting cultivation; Slash-and-burn; South East Asia; spatial data; Spectral feature-based algorithms; Swidden agriculture; systematic review; The tropics; South East Asia; Central Africa; Latin America; Slash-and-burn; The tropics; Swidden agriculture; Spectral feature-based algorithms; Dry season; Landsat
• ISSN: 1569-8432; 1872-826X; 1872-826X
• DOI: 10.1016/j.jag.2022.102876

•Three categories of remotely-sensed methods at different scales for swidden agriculture are reviewed.•Landsat has promoted the remote sensing of swidden agriculture since the 1970s.•Spectral feature-based algorithms dominate the remote sensing of swidden agriculture.•The slash and burn signature lays the foundation for remote sensing of swidden agriculture.•New data and algorithms are the directions for remote sensing of swidden agriculture. The evolution and transformations of swidden agriculture closely relate to ongoing studies of carbon budget, biological diversity, ecosystem services, indigenous wellbeing, and ethnic identities. Remote sensing has become an important tool to quantify and display swidden agriculture in transition. Due to its diversity, dynamics, and complexity, however, the development of universal remote sensing algorithms for mapping swidden agriculture faces major challenges. Although swidden agriculture has increasingly become the focus of climate change and other sustainable initiatives, a systematic review on remote sensing of this farming system in the tropics is seldom reported. With 89 peer-reviewed journal articles related to the identification, mapping, monitoring or forecast of swidden agriculture since the 1970s, the methodological progresses and shortcomings of remote sensing of swidden agriculture were compared and summarized, followed by providing several perspectives on near-future research directions. The conclusions were drawn as follows. Firstly, the limited remote sensing studies of swidden agriculture were mostly reported in Southeast Asia (66%), followed by Latin America and Central Africa. Secondly, Landsat is the most (59%) widely applied to monitor swidden agriculture and also holds huge potential in delineating its trajectories. Thirdly, the spectral feature-based algorithms (76%) are more used than the counterparts of geometric-based and structure-based ones. Finally, several aspects including exploring new data and algorithms for the enhancement of remote sensing of swidden agriculture have been proposed. This review paper might be served as a long-delayed appeal for strengthening remote sensing of swidden agriculture in the tropics for accurately assessing the carbon budget of forest change and other ecological and social effects caused by the evolution and transformations of swidden agriculture.