Estimating biophysical parameters in New Zealand grasslands: a scaling-up approach
Mapping biophysical parameters (LAI, nitrogen content, biomass, Green herbage ratio) is fundamental to understand ecosystem functioning. Considerable time and expense are commonly spent measuring these variables which can be scaled up from ecosystem to landscape level to monitor productivity, carbon stocks and fluxes. In a scaling up approach, remote sensing techniques can be used at ground, airborne and satellite level, and different Spectral Vegetation Indices (SVI) can be compared with data collected by direct harvesting.
Four main grassland classes (improved pastures, unimproved pastures, New Zealand native tussock grassland and depleted grasslands) were identified on Landsat ETM+ images. ASDI spectroradiometer and Landsat ETM+ data were examined at ground and satellite level (3x3 pixels), respectively. In January 2004, in New Zealand’s South Island, ground observations (Leaf Area Index, biomass, necromass and spectral measurements with a 170° diffusor optics) were performed in 11 sites belonging to the four grassland classes. Spectral signatures were determined at the two levels and NDVI was calculated for the different grassland canopies investigated.
The range of biomass levels within the different canopies was wide (from 10 to 667 g/m2 of dry matter). Necromass rate was high in tussock grasslands which also showed high levels of biomass. However, tussocks spectral signatures are similar to depleted areas.
NDVI calculated for improved pastures, unimproved pastures, and depleted grasslands is correlated with dry biomass levels both at ground (R2=0,76) and at Landsat ETM+ level (R2=0,74). Most of the tussock NDVI values do not follow a typically logarithmic trend. Saturation of NDVI occurs above biomasses of 100g/m2
The results of the present study confirm the potential of upscaling the biophysical parameters estimation from ecosystem to landscape level. Tussock spectral response is shown to be different from the other grasslands, because of the particular structure and the influence of bare soil.