Dissertation/Thesis Abstract

Separation of mixed radiometric land cover temperatures in time-delayed bi-angular views using estimated fractional differential coefficients
by Williams, Scott Lawrence, Ph.D., New Mexico State University, 2014, 201; 3582404
Abstract (Summary)

A dissertation is presented concerning the separation of radiometric temperatures of sparse land covers from two views of mixed thermal and NDVI samples with a time delay between the views. The research scope is limited to a simple binary land cover of vegetation canopy and bare soil. Previous methods have been developed using simultaneous views but little work has been done on time-delayed sampling, which is the focus of this study.

The dissertation hypothesis is based on the observation that the rate of change of a mixed radiometric temperature with respect to actual fractional vegetation cover, dTm/dfa originally constructed using spatially varying vegetation covers, can also be constructed using bi-angular views of the same land parcel but with a different interpretation; that bi-angular samples provide a perceived fractional cover differential, dTm/df0 . The hypothesis is that dTm/df0 can be used for sub-pixel temperature discrimination of binary land covers and, moreover, that the separate soil and vegetation total differential coefficients dTm/df0 and dTv/df0, required in the algebraic system, can be characterized to sufficiently capture environmental influences between samples in time. To test the hypothesis, this study heuristically derives a first-order estimation of the differential coefficients, required to decompose land cover temperatures from mixed data points, for any time-delayed sampling spanning the day. Applying the estimated values on similar target days gives a high success rate for a local time span of at least a week.

This approach, once scaled up, could be used by platforms with inherent time delays, such as tandem weather satellites, to provide separate land cover temperature estimates from low-resolution sensors.

Indexing (document details)
Advisor: DuBois, David
School: New Mexico State University
School Location: United States -- New Mexico
Source: DAI-B 76/07(E), Dissertation Abstracts International
Subjects: Agronomy, Electrical engineering, Environmental science
Keywords: Land Cover Temperatures, Remote Sensing, SOB-Pixel Analysis
Publication Number: 3582404
ISBN: 978-1-321-62061-0
Copyright © 2021 ProQuest LLC. All rights reserved. Terms and Conditions Privacy Policy Cookie Policy