Use this URL to cite or link to this record in EThOS:
Title: Retrieving forest characteristics from high-resolution airborne S-band radar data
Author: Ningthoujam, Ramesh Kumar
ISNI:       0000 0004 5918 8618
Awarding Body: University of Leicester
Current Institution: University of Leicester
Date of Award: 2016
Availability of Full Text:
Access from EThOS:
Access from Institution:
Synthetic Aperture Radar (SAR) data are utilized for improved mapping of forest cover and biophysical retrieval due to its sensitivity to forest canopy and structure. It is important to study the forest structure and biophysical parameters because it constitutes the major forest aboveground biomass (AGB). The S-band SAR frequency has not been consistently investigated for forest monitoring due to the lack of long-term data. Using the recent AirSAR campaign (2010-2014) over Savernake Forest and Wytham Woods in southern England, this thesis presents methods for analysing S-band SAR data for soil and forest canopies using the radiative transfer Michigan Microwave Canopy Scattering (MIMICS-I) model. The first result chapter shows that dominant scattering behaviour of S-band frequency arises from ground/trunk interactions with little direct crown scattering across all polarisations and incidence angles. The S-band backscatter shows significant sensitivity to both soil moisture content and surface roughness. Simulation experiments related to forest degradation show low co-polarisation backscatter due to reduced canopy component and tree density at S-band. Using the above information, the second result chapter shows that S-band HH- and VV-backscatter and Radar Forest Degradation Index (RFDI) data produces forest/non-forest classification map at 6 m resolution with 70% overall accuracy (kappa coefficient, κ = 0.41) while 63% overall accuracy (kappa coefficient, κ = 0.27) for the 20 m resolution map in a Maximum Likelihood algorithm. S-band data is also useful for mapping various non-forest cover types and monitoring forest cover changes over time due to the loss of volume scattering when forest canopies are removed. Using the field measured forest biomass, the third result chapter reveals that S-band radar backscatter correlates well with forest AGB. A consistent S-band backscatter/ biomass relationship is found, suggesting increasing backscatter sensitivity to forest AGB up to 100 t/ha with least error varying 90.46 - 98.65 t/ha at 25 m resolution (stand level) in low to medium incidence angles. The implications of these results are that S-band SAR data like the longer L-band SAR is highly suitable for mapping forest cover and monitoring cover changes and be able to retrieve low biomass stands below 100 t/ha.
Supervisor: Balzter, Heiko ; Tansey, Kevin Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available