An analysis of vegetation pattern and its relationship to NDVI data in the Namaqualand area, South Africa

The Namaqualand area in the North Western Cape, South Africa is unique in comparison to other similar semi-arid areas of the world. It has a high biodiversity and endemism and is consequently an area of interest for a growing number of conservation initiatives. Climate plays an important role in influencing the phenology and growth of the vegetation in the area. Remote sensing techniques were used to reveal the vegetation patterns in the greater Namaqualand area and to relate them to climatic variables. To do this we used the normalised difference vegetation index (NDVI) to relate biomass to altitude, rainfall and vegetation type. Each vegetation type in the area had a unique temporal signature and the climatic variables influencing the summer rainfall and winter rainfall vegetation types differed significantly from each other. Mean annual NDVI was significantly correlated to precipitation and potential evapotranspiration (PET) (r = 0.60, -0.63 respectively). A multiple regression model explained 52% of the variance when Mean Annual NDVI was related to climatic variables. Mean NDVI in August (the month of maximum NDVI in most of Namaqualand) was significantly related to PET and the current plus two previous months of precipitation (r = -0.72 and 0.74 respectively). A multiple regression model for mean NDVI in August and climatic variables explained almost 58% of the variance. The results suggest that NDVI can be used successfully as a measure of growth and phenology in the Namaqualand area and that NDVI could be used in climate models, drought prediction, desertification predictions and a number of other applications in the future.

Grantor dc:publisher.institution

Department of Biological Sciences

Year dc:date.issued

2003

Author dc:creator

• Fox, Sarah-Jane Caroline

Advisors dc:contributor.advisor

• Hoffmann, Timm
• Hoare, David

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/11427/25784

OAI identifier oai:identifier

oai:open.uct.ac.za:11427/25784