By Michele Hofmeyr
The information retrieved from satellite images taken from space often has to be checked and validated on the ground. This is exactly what the ambitious scientists in the Kruger National Park (KNP) are attempting to do.
Izak Smit, GIS and remote sensing researcher, and Nick Zambatis, vegetation ecologist from scientific services in Kruger, are currently working with Dr Konrad Wessels from the Meraka Institute, African Advanced Institute for Information and Communication Technology at the Council for Scientific and Industrial Research (CSIR) in Pretoria.
Dr Wessels is the research group leader for the remote sensing research unit of the Meraka Institute. “We are trying to relate the readings we are getting from the disc pasture meter, which measures how much herbaceous plant matter or biomass is on the ground, to the information we derive from satellite images of the same area,” explained Izak.
The team will most likely be using images from MODIS, a satellite sensor that passes over the park twice a day. This means two images of the entire Kruger Park and surrounding areas, taken with a spatial resolution that varies between 250m to 500m, are available on a daily basis. These images are freely available to be downloaded from the internet and are already used in Kruger for fire scar mapping.
Sanparks also receives email notification of active fires based on models developed for these satellite sensors. The team is hoping to establish a relationship between the information about vegetation that is visible from satellite images and what is actually growing on the ground.“We are trying to find out how and where to collect field data so that it can be successfully linked with the information from the satellite imagery.
It is hoped that we can eventually end up with a map that can be updated frequently for herbaceous biomass across the whole of Kruger, making use of the combined information gained from the field collected data and satellite imagery” said Izak.
In an initial trial run, the group has focused on open and wooded areas as well as low and high biomass areas, to get an idea of the variability in the landscape. The Rietpan firebreak south of Tshokwane, running from the well-wooded western granitic soils, across the shales and into the open, productive basaltic soils provided an excellent starting point.
Two scientific papers on this work have been published already, indicating that establishing this relationship between satellite images and the herbaceous layer is possible, but the team is hoping to devise a method that can be easily repeated and that is more operational. If this project is successful in estimating herbaceous biomass using satellite images, this will assist park ecologists as both a general vegetation monitoring and fire management tool.