Relationships between tropical, temperate and boreal forest variables and PALSAR data
Kevin Tansey(1), Heiko Balzter(1), Adrian Luckman(2), Agata Hoscilo(1) and Susan Page(1)
(1) University of Leicester, University Road, LE1 7RH, Leicester, United Kingdom
(2) Swansea University, Singleton Park, SA2 8PP, Swansea, United Kingdom
The overall aim of our work with ALOS PALSAR data is to develop methods to estimate forest variables (specifically biomass density and biomass density change) of forested regions in the UK (temperate forest), central Siberia (boreal forest) and Indonesia (tropical forest in Sumatra and Borneo). By utilising all of the operating modes of PALSAR, including the polarimetric mode and interferometric data from consecutive overpass, we explore the relationships between measured bio-physical forest variables (from field data) with values of backscatter coefficient, coherence and other data derived values. The paper will show our initial observations whilst addressing the following objectives:
1) Analyse the temporal variations in the polarimetric and the dual polarisation signal over an 18-month period for a selection of temperate, boreal and tropical tree stands.
2) Plot relationships between polarimetric scattering intensity and phase with physical attributes of forests (biomass, tree height, stocking volume etc.) to try and improve estimates of these parameters.
3) Explore the new observing technique known as polarimetric SAR interferometry or Pol-InSAR using 46-day repeat-pass data. Although temporal decorrelation of the coherence may be large there may be potential to retrieve phase information from low biomass regions and clear-cut regions, which is useful for determining forest extent).
The work will build on information gained from the E-SAR (2000) and ENVISAT (2001) missions.