Answers are grounded in this paper and cite their sources.
Land-cover change was evaluated for the Marahoué region of central-western Côte d’Ivoire (9,092.48 km²) using Landsat TM (1986), ETM+ (2002) and OLI (2019) imagery and a conditional spatialised Markov-chain projection to 2050. Five classes were mapped: water, settlements and bare soil, agricultural land, shrub savanna and dense forest. Field observations collected in 2017–2019 and a 2018 socio-economic survey supplied contextual information. Reported overall classification accuracies were 98.79%, 97.61% and 93.79%, with Kappa values of 0.98, 0.97 and 0.92 for 1986, 2002 and 2019, respectively; detailed validation matrices were not reported. Between 1986 and 2019, the mapped shares of settlements and bare soil increased from 12.49% to 18.58%, whereas the water class decreased from 19.23% to 15.64%. The modelled 2050 scenario assigns 25.74% of the region to settlements and bare soil and 38.40% to agriculture, compared with 18.58% and 30.11% in 2019. Dense forest, shrub savanna and water decline in this scenario. The retrospective 2019 comparison indicates a 2.01-percentage-point minimum redistribution between simulated and observed class totals but does not establish spatial prediction accuracy. The scenario therefore illustrates possible continuation of historical land-cover transitions rather than a climate-driven forecast. The findings are relevant to regional land-use monitoring, subject to uncertainties in classification and model validation.
No takes yet. Share an insight, caveat, or question.
Fabrice et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: