Remote sensing af vegetation

Kristine Bak Nielsen

Studenteropgave: Semesterprojekt

Abstrakt

This report concerns the correlation between the two vegetation indices NDVI and EVI obtained from satellite measurements and observed data for the gross primary production in Lille Bøgeskov near Sorø, Zealand, Denmark. The vegetation indices are based on the ratio between reflection of visible and near-infrared light. Plants will absorbe visible light in the process of photosynthesis and reflect near-infrared light. Because of this indices based on these wavelengths will be sensitive towards vegetation and can be used to measure the density of vegetation. Mean values for NDVI and EVI from EOS/MODIS has been calculated with ArcGIS over Lille Bøgeskov in 2010. The vegetation indices have been filtered for days with snow and days with limited usefullness. The mean values have been compared with leaf area index measured by an optic sensor LAI-2000 and with leaf area index estimated from photosynthetic active radiation using Beer’s law. Beer’s law contains the extinction coefficient k that depends on the vegetation conditions. The extinction coefficient will vary over a year and has not been estimated. There is a high correlation between the filtered vegetation indices and the leaf area index estimated from photosynthetic active radiation (R2 = 0.94). There is no big difference between the cor- relation of NDVI or EVI. NDVI seems to show an asymptotic behaviour towards increasing leaf area index while EVI seems to have a more linear relation to leaf area index.

UddannelserGeografi, (Bachelor/kandidatuddannelse) Bachelor el. kandidat
SprogDansk
Udgivelsesdato16 jun. 2011

Emneord

  • GIS
  • vegetation indices
  • leaf area index
  • MODIS