Remote sensing applied in land use and land cover change (LUCC) in arid and semi-arid ecosystems : current status, challenges and prospects – a systematic review

Abstract

Arid and semi-arid ecosystems face a triple environmental threat at the intersection of climate change, biodiversity loss, and desertification. Land use and land cover change (LUCC) in the region is, therefore, a key indicator in monitoring overall ecosystem health. However, persistent challenges hinder LUCC assessment in dry biomes, notably due to the heterogeneity of arid vegetation and its response to high temperature and water availability variations throughout the year. To identify the most important knowledge gaps easily tackled in the near future, this systematic review synthesizes current scientific knowledge on LUCC monitoring in arid and semi-arid ecosystems from the Scopus and Dimensions databases. Study results showed that seasonality (41%) and spatial resolution (28%) were the two main factors limiting LUCC in arid and semi-arid ecosystems. These factors had a significant influence on map accuracy, particularly due to seasonal variations in vegetation and the conditions under which satellite images were taken. Classification accuracy decreased substantially when the studies attempted to describe vegetation classes in more detail. Reported classification accuracies in the literature (accuracy ≥ 80%) reflected overconfidence, as a large proportion of the publications (70%) were limited to general distinctions such as forest and non-forest, without tackling more complex categories. Thus, given the importance of global vegetation maps for ecosystem management, a valuable tool for forest management and environmental planning, wrong estimation can introduce biases, leading to inappropriate management decisions and thus compromising the sustainability of natural resources and the ecosystem services that depend on them.