Oct 29 – 31, 2026
Essaouira, Morocco
Africa/Casablanca timezone

Coastal Flood Susceptibility Mapping in Essaouira, Morocco: An Integrated GIS and Machine Learning Approach

Oct 29, 2026, 2:45 PM
15m
ARGANA (Municipal Library of Essaouira)

ARGANA

Municipal Library of Essaouira

Speaker

Ms Chaimaa EL KIHAL (Cadi Ayyad University (UCA), Essaouira Higher School of Technology, Interdisciplinary Laboratory for Research in Environment, Management, Energy and Tourism (LIREMET), Av Abdelkrim Khattabi, B.P. 511 - 40000, Marrakech, Morocco.)

Description

Climate change, rapid urbanization, and sea-level rise are increasingly intensifying flood susceptibility in coastal areas such as Essaouira, Morocco. This study develops an integrated GIS, remote sensing, and machine learning framework to identify flood-prone zones and support evidence-based flood risk management. A cleaned dataset of 1,529 georeferenced flood and non-flood points was prepared and divided into training, validation, and testing subsets. The analysis incorporated key flood-conditioning factors representing topographic, hydrological, climatic, coastal, and land-use conditions, including elevation, slope, aspect, land use/land cover, TWI, TPI, SPI, plan and profile curvature, distance to rivers, distance from the coastline, precipitation, and temperature. Two machine learning models were applied and compared: Random Forest and Artificial Neural Network. Random Forest achieved the highest predictive performance, with a test accuracy of 95.1% and an AUC of 0.983, demonstrating its strong capacity to model nonlinear relationships and identify the most influential flood-related variables. The ANN model also showed robust performance, reaching 92.0% accuracy and an AUC of 0.962, confirming its ability to capture complex spatial interactions between environmental and climatic factors. The resulting susceptibility maps indicate that the most flood-prone areas are mainly concentrated in low-lying coastal plains, zones close to rivers and temporary wadis, and rapidly urbanizing sectors where impervious surfaces increase runoff and reduce infiltration. These findings reflect the local conditions of Essaouira, where coastal exposure, low elevation, rainfall variability, drainage characteristics, temperature-related climatic conditions, and land-use changes collectively contribute to flood vulnerability. Overall, this study demonstrates that combining GIS, remote sensing, RF, and ANN provides a reliable and transferable decision-support approach for flood susceptibility mapping, urban planning, disaster risk reduction, and climate adaptation in vulnerable coastal environments.
Keywords: Flood susceptibility, Artificial Intelligence, GIS, Machine learning, Remote Sensing, Essaouira.

Authors

Ms Chaimaa EL KIHAL (Cadi Ayyad University (UCA), Essaouira Higher School of Technology, Interdisciplinary Laboratory for Research in Environment, Management, Energy and Tourism (LIREMET), Av Abdelkrim Khattabi, B.P. 511 - 40000, Marrakech, Morocco.) Mr Mohamed IKANBANE (Cadi Ayyad University (UCA), Essaouira Higher School of Technology, Interdisciplinary Laboratory for Research in Environment, Management, Energy and Tourism (LIREMET), Av Abdelkrim Khattabi, B.P. 511 - 40000, Marrakech, Morocco.) Kholoud KAHIME (Cadi Ayyad University (UCA), Essaouira Higher School of Technology, Interdisciplinary Laboratory for Research in Environment, Management, Energy and Tourism (LIREMET), Av Abdelkrim Khattabi, B.P. 511 - 40000, Marrakech, Morocco)

Presentation materials

There are no materials yet.