Document Type

Dissertation - Open Access

Award Date

2019

Degree Name

Doctor of Philosophy (PhD)

Department / School

Electrical Engineering and Computer Science

First Advisor

Dennis Helder

Second Advisor

Larry M. Leigh

Keywords

Classification of Landcover, Cross-calibration, Extended Pseudo Invariant Calibration Sites, Hyperspectral, Multispectral, Landsat 8, longterm trending of sensor

Abstract

An increasing number of Earth-observing satellite sensors are being launched to meet the insatiable demand for timely and accurate data to help the understanding of the Earth’s complex systems and to monitor significant changes to them. The quality of data recorded by these sensors is a primary concern, as it critically depends on accurate radiometric calibration for each sensor. Pseudo Invariant Calibration Sites (PICS) have been extensively used for radiometric calibration and temporal stability monitoring of optical satellite sensors. Due to limited knowledge about the radiometric stability of North Africa, only a limited number of sites in the region are used for this purpose. This work presents an automated approach to classify North Africa for its potential use as an extended PICS (EPICS) covering vast portions of the continent. An unsupervised classification algorithm identified 19 “clusters” representing distinct land surface types; three clusters were identified with spatial uncertainties within approximately 5% in the shorter wavelength bands and 3% in the longer wavelength bands. A key advantage of the cluster approach is that large numbers of pixels are aggregated into contiguous homogeneous regions sufficiently distributed across the continent to allow multiple imaging opportunities per day, as opposed to imaging a typical PICS once during the sensor’s revisit period. In addition, this work proposes a technique to generate a representative hyperspectral profile for these clusters, as the hyperspectral profile of these identified clusters are mandatory in order to utilize them for performing cross-calibration of optical satellite sensors. The technique was used to generate the profile for the cluster containing the largest number of aggregated pixels. The resulting profile was found to have temporal uncertainties within 5% across all the spectral regions. Overall, this technique shows great potential for generation of representative hyperspectral profiles for any North African cluster, which could allow the use of the entire North Africa Saharan region as an extended PICS (EPICS) dataset for sensor cross-calibration. Furthermore, this work investigates the performance of extended pseudo-invariant calibration sites (EPICS) in cross-calibration for one of Shrestha’s clusters, Cluster 13, by comparing its results to those obtained from a traditional PICS-based cross-calibration. The use of EPICS clusters can significantly increase the number of cross-calibration opportunities within a much shorter time period. The cross-calibration gain ratio estimated using a cluster-based approach had a similar accuracy to the cross-calibration gain derived from region of interest (ROI)-based approaches. The cluster-based cross-calibration gain ratio is consistent within approximately 2% of the ROI-based cross-calibration gain ratio for all bands except for the coastal and shortwave-infrared (SWIR) 2 bands. These results show that image data from any region within Cluster 13 can be used for sensor crosscalibration. Eventually, North Africa can be used a continental scale PICS.

Library of Congress Subject Headings

Landsat satellites -- Calibration -- Africa, North.
Artificial satellites in remote sensing -- Calibration -- Africa, North.
Imaging systems -- Image quality.
Remote sensing -- Data processing.

Description

Includes Creative Commons Atrribution 4.0 materials

Format

application/pdf

Number of Pages

99

Publisher

South Dakota State University

Share

COinS
 

Rights Statement

In Copyright