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

Accueil du site → Doctorat → États-Unis → 2013 → Remote Sensing Tools For Land And Water Management In Data Scarce Blue Nile Basin

Cornell University (2013)

Remote Sensing Tools For Land And Water Management In Data Scarce Blue Nile Basin

Ayana, Essayas Kaba

Titre : Remote Sensing Tools For Land And Water Management In Data Scarce Blue Nile Basin

Auteur : Ayana, Essayas Kaba

Université de soutenance : Cornell University

Grade : Doctor of Philosophy (PhD) 2013

Ground based water resources monitoring systems are often difficult to maintain consistently in developing countries. The decline in the number of stations, data quality and changes in the data holding policy has made water resources data less reliable for use in operational purposes. The objective of this dissertation is, therefore, to evaluate the utility of existing freely available remotely sensed images to monitor water resource systems. In this dissertation Moderate Resolution Imaging Spectroradiometer (MODIS) images were evaluated on the basis of their capability to (1) measure total suspended solid (TSS) and turbidity and generate historical TSS data, (2) estimate the water storage variation of Lake Tana and (3) monitor the state of biomass in the upper Blue Nile basin . The usability of historical TSS data in hydrologic modeling is also tested. Lake water samples were collected concurrent with the satellite overpass over the lake at the entry location of Gumera River, a major tributary to the lake. Reflectance in the red and near infrared (NIR) 250 m-pixel images taken on sampling days were correlated and validated using measured TSS and turbidity. The validated correlations were applied to the ten year image archive of MODIS to generate a 10-year TSS time series for the lake. In addition, MODIS images of the years 2002 - 2003, where the lake level variation was at its minimum, were used to generate the lake near-shore bathymetric model. The new near-shore bathymetric model reproduced water level measurements with a better accuracy than the existing bathymetric model of the lake. The usability of the TSS data was tested by initializing a hydrologic model for the Gumera watershed using the Soil and Water Assessment Tool (SWAT). The ten year TSS data generated were used to calibrate the model. The model was capable of predicting the monthly TSS variation. The potential of MODIS images in monitoring biomass recovery was also assessed at river basin scale. The enhanced vegetation index (EVI) - land surface temperature (LST) relation is used to map the trend in the disturbance of plantations put in place as conservation measures. In this dissertation the potential of satellite imagery as a data gap filling alternative to ground based monitoring systems in data scarce regions is tested.


Version intégrale (4,38 Mb)

Page publiée le 19 février 2016, mise à jour le 7 octobre 2017