Use this URL to cite or link to this record in EThOS: https://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.755358
Title: Forest extent monitoring in the Congo Basin using L-band Synthetic Aperture Radar (SAR)
Author: Wheeler, James Edward Maxwell
ISNI:       0000 0004 7428 3533
Awarding Body: University of Leicester
Current Institution: University of Leicester
Date of Award: 2018
Availability of Full Text:
Access from EThOS:
Access from Institution:
Abstract:
This thesis presents improved forest extent monitoring using L-band Synthetic Aperture Radar (SAR) data freely provided by the Japanese Aerospace Exploration Agency (JAXA) over the Congo Basin, which contains the second largest area of rainforest in the world. Forest loss in the region, estimated to be up to 0.63 billion Mg in the period 1990-2005, is predominantly driven by bush fuel collection, at a characteristically small scale. Single medium resolution SAR scene (75 km x 75 km), wide area (550 km x 550 km) and full Central African (2000 km x 3300 km) forest extent classifications are generated, and inform best practice for operational annual forest cover production from L-band SAR data. Improvements in one or more of overall accuracy, consistency, scope and replicability are observed compared with existing wide area forest cover and forest/non-forest products in Central Africa, using robust statistical methods to quantify errors in reported class areas. Seasonally inundated forest, a regional obstacle to previous SAR forest cover classifications, is identified by the range of co-polarised SAR data and tested using a novel metric incorporating a normalised cumulative rainfall value aggregated by sub-basin catchment area and SAR polarimetric analysis, which is itself compared with coarse resolution Soil Moisture and Vegetation Water Content Metrics from the Advanced Microwave Scanning Radiometer for the Earth Observing System (AMSR-E). A quantitative analysis of multiple SAR resolutions supports the continued and future use of 30 m and higher resolution L-band SAR data to map forest cover in the region.
Supervisor: Tansey, Kevin ; Balzter, Heiko Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.755358  DOI: Not available
Share: