|
|
|
 |
|
|
Presentation
< back to presentation overview
| Details of the presentation |
| Presentation |
Oral presentation |
| Title |
Using Remote Sensing And Expert Knowledge To Map Landscape-Level Land Degradation In South Africa And Namibia
|
| PDF Download |
--- |
| Short title |
Remote Sensing to Map Land Degradation
|
| Author(s) |
Wistebaar, N.(1); Hoffman, T.(2); Desmet, P.(3); Rouget, M.(1); Jonas, Z.(1); Lieckfeld, L.(4); Gessner, U.(5)
|
| Presenting author |
Th. Wistebaar; Rouget, M.
|
| Institution(s) |
(1) South African National Biodiversity Institute, Brummeria; (2) Plant Conservation Unit, Botany Department, University of Cape Town; (3) 84 Clearwater Street, Glenwood, Pretoria; (4) German Aerospace Center (DLR), German Remote Sensing Data Center, Wessling; (5) Remote Sensing Unit in cooperation with DLR, University of Wuerzburg
|
| Keywords |
Dryland degradation, overgrazing, semi arid rangelands, remote sensing, MODIS, NDVI, phenometrics
|
| Abstract |
We present different studies investigating the use of remote sensing data to quantify land degradation based on spatial and temporal scales. In dryland areas, both high inter annual rainfall variation and high spatial variability of land conditions determine vegetation growth and aggravate land degradation. As a result, quantify changes due to anthropogenic pressures is difficult.
Within the Bushmanland region in NW South Africa, ANOVA test per vegetation response phase were carried out to establish the phenological variable which best detects change in the vegetation cover. The NDVImin, NDVImax and NDVI∑ variables were found to best explain differences between the three vegetation units. Taking into account rainfall, vegetation units and phenology, two spatial and temporal approaches were developed which used NDVI to quantify land degradation. The spatial scale approach, based on the dry phase only, used the benchmark method to establish thresholds for any changes in veld condition. The temporal scale approach used the residual method based on seven year averaged NDVI. Regression and mean analysis were carried out based on the residual values for each sample point. Overall, despite a lack of appropriate ground-truthing data, the two derived methods have implications to spatially and temporally quantify land degradation in arid and semi-arid environments.
In regional studies in Namibia, the effects of different farming systems and stocking rates on savanna ecosystems are reflected in specific vegetation classes. In this context it was important to have detailed information on the study area, and to include expert knowledge to correctly interpret the remotely sensed vegetation patterns. Detailed botanical field surveys and farmer interviews on the different farming systems were used in combination with remote sensing analyses to identify degradation due to human impact. Further refinement of the methodology is necessary, including ground-thruthing for validation purposes.
The presented methodologies showed promise for monitoring and mapping grazing carrying capacity and overgrazed areas. An essential issue of the results is the use of rainfall data, expert knowledge and other ground truth information for differentiating between natural variations and degradation due to anthropogenic pressures.
|
| Congress Topic |
Observation System
|
| Topic No. |
1.3 |
| Notes |
---
|
| Ref. No. |
604 |
|
|
|
