Izzawati; (2002) The use of a radar backscatter model to assess the sensitivity of multi-frequency, -polarisation, and -angle SAR data for detecting the growth stage of oil-palm plantations. Doctoral thesis (Ph.D), UCL (University College London).

Text The_use_of_a_radar_backscatter.pdf Download (25MB)

Abstract

Increasing interest in SAR data for forest monitoring has encouraged the use of SAR sensors with varying frequency, polarisation and incidence angle. As microwave signals of different frequencies and polarisations interact with different parts of tree canopies in a characteristic manner, a physical model can assist in understanding how tree growth and canopy changes affect microwave back scattering signatures. The model can be used to conduct sensitivity analysis to these changes, and this sensitivity assessment will enable the recommendation of optimum radar parameters to be employed, thus improving the effectiveness of such a monitoring task. This research has adapted and employed a three-dimensional radar backscatter model previously developed by Sun and Ranson (1995) to assess the sensitivity of ERS and JERS SAR data for detecting the growth stage of oil-palm plantations, In performing this assessment, the model is run at various oil-palm tree growth stages to assess the sensitivity of different radar configurations. Results from this sensitivity analysis show that ERS CVV and JERS LHH SAR configurations are not optimal for such task. While ERS CVV performs better than JERS LHH for detecting the growth stage of Oil-Palm plantation at 4, 8 and 12 years old, the optimum single frequency/polarisation combination is shown to be CHV. Results from the simulation also show that the sensitivities of X-, C- and L-band are significantly affected by the condition of the ground surface underneath the plantation. As the ground condition can vary significantly, from swamp to understory and plantation litter, any attempt to differentiate the growth stage of such plantations will be very difficult without prior knowledge of the ground conditions. Simulation results suggest that the use of multi-frequency, -polarisation and -angle data, however, enable the differentiation of the major scattering mechanisms and a characterisation of the ground condition. Using this information, simple methods to differentiate the ground condition and the plantation growth stages at various ground surface condition are proposed.

Download activity - last month

Export as JSONCSVXML

Download activity - last 12 months

Export as JSONCSVXML

Downloads by country - last 12 months

Export as CSVXMLJSON

Archive Staff Only

View Item