Hydrological modelling in a data-scarce semi-arid catchment: Assessing urbanisation impacts on runoff in Bouskoura, Morocco

Abstract

Urbanisation is a major driver of hydrological change, particularly in semi-arid regions where data scarcity limits the calibration of hydrological models. This study addresses the challenge of modelling hydrological responses in the ungauged Bouskoura catchment (Morocco), which has experienced rapid urban growth over recent decades. We combined land-use maps from 2017 and 2021, developed from Sentinel imagery and classified with Support Vector Machine (SVM) algorithms (κ = 0.99), with demographic projections to 2030 and applied the HEC-HMS hydrological model with the SCS-CN method to assess changes in runoff dynamics. Results show that the runoff coefficient increased from 69.11 in 2017 to 72.06 in 2021, with projections reaching 77.56 by 2030. Peak discharges followed the same trend, rising from 28.4 m³∙s⁻¹ in 2017 to 42.1 m³∙s⁻¹ in 2021 and up to 92.9 m³∙s⁻¹ in 2030, indicating an alarming escalation of flood risk under continued urban expansion. While the absence of observed streamflow data prevents calibration, the comparative scenario-based framework applied here provides reproducible insights into hydrological shifts in data-scarce semi-arid basins. These findings highlight the urgent need to integrate stormwater management and sustainable land-use planning in peri-urban areas of Casablanca and, more broadly, demonstrate the applicability of remote sensing and model-based approaches for ungauged catchments facing rapid urbanisation.