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

Accueil du site → Doctorat → Royaume-Uni → 2002 → Use of high resolution remote sensing and GIS to parameterise spatially-distributed soil erosion models (South-east Spain)

University of London - King’s College London (2002)

Use of high resolution remote sensing and GIS to parameterise spatially-distributed soil erosion models (South-east Spain)

Nadeem Hashem

Titre : Use of high resolution remote sensing and GIS to parameterise spatially-distributed soil erosion models (South-east Spain)

Auteur : Nadeem Hashem

Université de soutenance : University of London - King’s College London

Grade : Doctor of Philosophy (PhD) 2002

Résumé
South-east Spain is subject to soil erosion and land degradation as a result of natural and human-induced factors such as the change in landuse pattern, coupled with a semi-arid climate. Estimates of soil erosion are needed by decision makers in order to quantify the soil loss under various scenarios and different landuse patterns so that policies of sustainable use of land resources can be developed and implemented. The purpose of this study is to explore the potential of using high resolution aerial photography and GIS to derive the parameters controlling the soil loss spatially at the local scale and then implement and validate a soil-erosion model. This enables the automation of the process of soil-erosion modelling to make it cost effective and less labour intensive. The Thornes model was selected for soil-emsion modelling in this study because it can be implemented in a spatially distributed manner and it needs little data to parwrieterise. Four factors control the soil loss in the Thornes equation : slope, vegetation cover, soil erodibility, and overland-flow. A Digital Elevation Model (DEM) was derived from the aerial photographs using digital photogrammetry techniques and slope was then calculated from the DEM. Various methods of classification and vegetation indices were reviewed to map the vegetation cover and the mixture modelling method was implemented because in this method the shade can be mapped and consequently removed from the vegetation map. A co-occuffence matrix and image texture analysis was used to correct the vegetation errors in a few pixels. Soil erodibility was calculated using Wischmeier’s equation. The Carson and Kirkby model was used to estimate the overland flow because it war, shown by other studies to produce satisfactory results in the ttudy area. Detenninistic and stochastic approaches were used for overland-flow modelling and the first approach produced results of higher accuracy and, thus, was implemented. rfbe parameters controlling the soil loss were then integrated according to ’Thornes equation in a GIS environment producing the soil loss map. This map was validated against measured values resulted from rainfall simulation experiments and the accuracy was satisfactory. Tk effect of landuse change by ploughing on soil loss was assessed and it was shown that ploughing may increase soil loss by up to 500%.

Présentation

Version intégrale (68 Mb)

Page publiée le 30 décembre 2019