Use this URL to cite or link to this record in EThOS: http://ethos.bl.uk/OrderDetails.do?uin=uk.bl.ethos.637270
Title: Deriving albedo and biophysical properties from coarse spatial resolution multiple-view-angle data
Author: Hesley, Z. J.
Awarding Body: University of Wales Swansea
Current Institution: Swansea University
Date of Award: 1999
Availability of Full Text:
Access from EThOS:
Abstract:
Previous investigations have focused on the development of bidirectional reflectance distribution function (BRDF) models and the implementation of computationally-efficient procedures for model inversion. The present study explores the operational use of BRDF models since little is known about producing BRDF at an operational level. Earlier work indicated that the angular sampling regimes of sensors influenced the accuracy of albedo estimated from BRDF models and multiple view angle (MVA) data. This work takes this a stage further by attempting to quantify the albedo error from present and future sensors, like SPOT VEGETATION, ATSR2, MODIS and MISR. The simulations demonstrated that the solar zenith angles (SZA) and the number of samples influences the accuracy of albedo derived. These findings suggest albedo can be accurately derived from the sensors studied. Contemporary work has shown the problem of residual mis-registration between images acquired by fine resolution MVA data, but this work shows new advances by examining mis-registration for coarse resolution AVHRR data. The results, even at these coarse scales, show similar mis-registration sensitivity curves to previous work. The implications of this result are that images needed to produce operational BRDF products will need sub-pixel registration accuracy and that mis-registration effects could swamp out the directional component of MVA data. Inversion of BRDF models against remotely sensed data has been used to derive biophysical parameters. Little is known, however, about the issues of mixed pixels and the derivation of biophysical parameters. A linear mixture model (LMM) is used in this study to unmix pixels before model inversion. The findings suggest the unmixing of pixels could improve the quality of BRDF products produced by sensors.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID: uk.bl.ethos.637270  DOI: Not available
Share: