Science Service System
You are here : Home : Proposals_Summary

Summary of Proposal LAN1637

TitleRemote Sensing Mapping of Arid Zone Vegetation
Investigator Bruce, David - University of South Australia, Barbara Hardy Institute
Team Member
Mr Alsharrah, Saad - University of South Australia, Barbara Hardy Institute
Dr Somenahalli, Sekhar - University of South Australia, Barbara Hardy Institute
Professor Bouabid, Rachid - National Agricultural School of Meknes, Department of Soil Sciences
SummaryThis research project focuses on mapping vegetation cover in a desert environment located in the Meknes-Tafilalet region in Morocco.The scope of the investigation involves acquiring high resolution radar (TerraSAR-X) and optical VNIR (Pleiades) imagery for accurate mapping of arid zone vegetation.Vegetation information from remote sensing is now being used in desertification modelling and assessment, which aids in developing actions plans to combat desertification.Arid environments typically have sparsely vegetated areas which makes it more difficult to retrieve quantitative information on vegetation type, cover and biomass because of the dominance of soil background reflectance.In arid environments, vegetation cover is light and varies significantly over space and time.In addition, the soil background in arid environments is highly variable and therefore sensitivity at low leaf area is limited in VNIR vegetation indices.Some challenges exist in mapping and discrimination of surface features related to land degradation.One is the spectral similarity amid surface components related with land degradation.Interpretations of remotely sensed measurements of sparse vegetation is also a difficult task and attempts to relate vegetation indices to fractional vegetation cover information are somewhat questionable.Additionally, vegetation such as small shrubs tends to be smaller than the imagery pixel size and is hard to map using classical methods.This research aims to utilise TerraSAR-X imagery to extract information on vegetation structure in an arid environment.The dual polarization capabilities of the HighResolutionSpotLight mode will be used extract structural information on low level desert shrubs and plants.The required data are two dual polarized HighResolution SpotLight (HS) images for the study site: 1) August 2012 (Archived) or as soon as possible, and 2) a second acquisition in September 2013. Additionally, high resolution optical imagery will be utilised to assist in mapping vegetation cover.It is noted that optical satellite sensors incorporate limitations in distinguishing between vegetation types where utilisation of RS techniques that are based on solely on the spectral information of the VNIR region will find difficulties it vegetation discrimination given the saturation of the observed signal with increasing biomass.Hence radar imagery, with its ability to extract textural information on the target object, can provide useful information for distinguishing between vegetation types based on their geometric properties.Both data sets will be processed and fused using two methods:1) processing each data set and estimating parameters, followed by combining both data sets to estimate vegetation cover for the study site, and 2) utilise a physical model to combine the data for single-parameter estimation.Funding for this research project will mostly be provided from Saad Alsharrah’s PhD fund and the awarded International Travel Grant by the University of South Australia.

Back to list of proposals

© DLR 2004-2016