Informations et ressources scientifiques
sur le développement des zones arides et semi-arides

Accueil du site → Master → Etats Unis → 2019 → APPLICATION OF A DISTRIBUTED HYDROLOGIC MODEL FOR THE ANALYSIS OF LAND USE CHANGE IN KEDOUGOU, SENEGAL

Michigan Technological University (2019)

APPLICATION OF A DISTRIBUTED HYDROLOGIC MODEL FOR THE ANALYSIS OF LAND USE CHANGE IN KEDOUGOU, SENEGAL

Perez, Francisco,

Titre : APPLICATION OF A DISTRIBUTED HYDROLOGIC MODEL FOR THE ANALYSIS OF LAND USE CHANGE IN KEDOUGOU, SENEGAL

Auteur : Perez, Francisco,

Université de soutenance : Michigan Technological University

Grade : Master of Science in Geology (MS) 2019

Résumé
West Africa has experienced extensive land use/land cover (LULC) change since the Sahel drought of the 1970s and 80s. Demographic changes, rapid urbanization, and other anthropogenic impacts on land cover in West Africa have potentially altered hydrological regimes. This work was undertaken while I served as a Peace Corps Volunteer in the Kedougou region of Senegal, an area that has experienced significant and rapid infrastructural improvements and urbanization since the early 2000s. The Gambia River flows near Kedougou, where flooding is an extensive problem, making access and livelihood practices difficult during the rainy season. Changes in climate and land use on hydrology have not been adequately assessed for this region, as well as many underdeveloped areas. A computational hydrological analysis is needed to understand recent and current conditions and forecast future impacts. The physically based and distributed Gridded Surface/Subsurface Hydrologic Analysis (GSSHA) model was employed to assess the effects of LULC change on inundation and local hydrology between 1992 and 2015. Urban area comprised 2.75 km2 of a 2,488-km2 watershed in 2015, growing from nothing in 1992. Forest cover between 1992 and 2015 increased 6%, whereas shrubland decreased by 4%. The 2015 land cover results showed a 20% increase in peak discharge during high intensity precipitation events (≥100 m). However, during lower intensity storms, urbanization did not significantly impact local hydrology or streamflow. In fact, a reverse trend was observed, in which peak flow was 25 m3/s higher during the 1992 land cover scenario. Increased runoff may have resulted from the vegetation loss influenced by the drought from the 1980s. When compared to Sentinel-1 SAR imagery, GSSHA moderately overpredicted inundation extent.

Présentation

Version intégrale (3,9 Mb)

Page publiée le 19 décembre 2019