Land condition monitoring : integrating ground surveys, Aussie GRASS and satellite remote sensing

Abstract

The Land Condition Monitoring Program (LCMP) was established in 1998 to assess a range of management issues across the agricultural region of South Australia. The potential for integrating ground data of the LCMP’s windscreen survey, satellite remote sensing and Aussie GRASS products was identified by South Australia’s Soil Conservation Council. Satellite derived NDVI has been used successfully in the past to monitor green vegetation. The Aussie GRASS model is also currently used to monitor growth of forage in the rangelands of Australia. A relationship was found between the LCMP cropping land survey, Aussie GRASS products Total Pasture Growth, Total Standing Dry Matter and rainfall, and satellite derived NOAA NDVI. Between 70% and 80% of cover can be explained by Aussie grass products TPG, TSDM and rainfall. As the Aussie GRASS model is dependent upon climatic factors, it can also be said that 70-80% of cover can therefore also be explained by climate. The remaining 20-30% of cover could be explained by management practices, which are considered when ranking cover in the windscreen survey. However, this error variance may also be a result of Aussie GRASS not sufficiently calibrated in the cropping regions of South Australia. With a relationship found between Aussie GRASS and the windscreen survey, cover was modelled for the entire cropping region. This allows for a continuos visualization of cover across State, which is not achievable with survey results alone. Analysis of data much similar to that achieved with the windscreen survey can also be performed on this data, but instead of grouping by zones, grids of 5km pixel size, which provide a finer view of cover, can be used. With cover modelled by Aussie GRASS products, this research project explored the possibilities of rating detachment by land use and season in order to create continuous maps of EHI. At present more work needs to be done on a accurate method of estimating detachment that takes into account different land management practices across the agricultural region. Wind and water erosion potential maps from the DWLBC Soil and Land Information soil database were investigated for their use in assessing erosion risk from the EHI. These maps are a better depiction of landscape features than those observed in the windscreen survey. No relationship was found between NPLR survey data, Aussie GRASS products and satellite remotely sensed NDVI. The NPLR was not included in field surveys used to calibrate the Aussie GRASS model. It is therefore possible that no relationship was found between the LCMP survey data and Aussie GRASSS products, as the model has not been calibrated in the NPLR.