A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean
The research was carried out with a view to developing a procedure for the computation of sea surface advection velocities from pairs of NOAA AVHRR infrared images. The procedure was designed for application to the oceanic regions around South Africa and cognisance had to be taken of restrictions im...
Main Author: | |
---|---|
Other Authors: | |
Format: | Doctoral Thesis |
Language: | English |
Published: |
University of Cape Town
2017
|
Subjects: | |
Online Access: | http://hdl.handle.net/11427/23143 |
id |
ndltd-netd.ac.za-oai-union.ndltd.org-uct-oai-localhost-11427-23143 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-netd.ac.za-oai-union.ndltd.org-uct-oai-localhost-11427-231432020-07-22T05:08:03Z A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean Agenbag, Johannes Jacobus Shillington, Frank A Oceanography The research was carried out with a view to developing a procedure for the computation of sea surface advection velocities from pairs of NOAA AVHRR infrared images. The procedure was designed for application to the oceanic regions around South Africa and cognisance had to be taken of restrictions imposed by the specific oceanographic conditions, availability of satellite data, as well as the capabilities of the image processing system used. As a first step, a set of image navigation algorithms was developed, based on elliptical orbit and ellipsoidal earth models. Orbit parameters were obtained from TBUS-bulletins and one or more ground reference points had to be identified on each. The navigation algorithms were then used to develop a procedure for the geometric transformation of images to a Mercator map projection. The transformation procedure was evaluated through use of test-images and the results indicated that the maximum errors which could be expected in the computation of advection vectors were 4-5 cm/sin the north/south velocity component and 6-7 cm/sin the east/west component if two images, 12 hours apart in time, were used for the vector computation. An automatic feature tracking method was tested as a means for computing advection velocities but was found to be unsatisfactory. As a result, a 'semi-automated' procedure was developed. This process is essentially a manual (point-wise) feature tracking procedure into which the template matching technique which formed the basis of automated procedures, was incorporated as a labour saving device. Tests indicated a time saving of 20-40 % on the manual procedure and more rapid computation than with the automated procedure. The feature tracking procedure was applied to three sets of AVHRR images of the South East Atlantic. To assess the precision of the vector computation procedure, two independent vector sets were computed. A comparison of the two sets indicated that the rootmean- square deviation in vector magnitude (speed) was about 6-8 cm/sand in the vector direction, about 31° (12° if very small vectors ≤ 6 cm/s are excluded). The computed vectors compared very well with reported results from conventional methods. The derived vector fields also provide the first really detailed description of surface currents in the sea off South Africa: eg. on the flow field in the southern Benguela Current, the circulation associated with Agulhas Current rings, and advective influences on the transport of fish eggs and larvae from the spawning grounds on the Agulhas Bank to the favoured recruitment area off the West Coast. 2017-01-26T06:44:43Z 2017-01-26T06:44:43Z 1992 2016-11-22T09:01:54Z Doctoral Thesis Doctoral PhD http://hdl.handle.net/11427/23143 eng application/pdf University of Cape Town Faculty of Science Department of Oceanography |
collection |
NDLTD |
language |
English |
format |
Doctoral Thesis |
sources |
NDLTD |
topic |
Oceanography |
spellingShingle |
Oceanography Agenbag, Johannes Jacobus A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean |
description |
The research was carried out with a view to developing a procedure for the computation of sea surface advection velocities from pairs of NOAA AVHRR infrared images. The procedure was designed for application to the oceanic regions around South Africa and cognisance had to be taken of restrictions imposed by the specific oceanographic conditions, availability of satellite data, as well as the capabilities of the image processing system used. As a first step, a set of image navigation algorithms was developed, based on elliptical orbit and ellipsoidal earth models. Orbit parameters were obtained from TBUS-bulletins and one or more ground reference points had to be identified on each. The navigation algorithms were then used to develop a procedure for the geometric transformation of images to a Mercator map projection. The transformation procedure was evaluated through use of test-images and the results indicated that the maximum errors which could be expected in the computation of advection vectors were 4-5 cm/sin the north/south velocity component and 6-7 cm/sin the east/west component if two images, 12 hours apart in time, were used for the vector computation. An automatic feature tracking method was tested as a means for computing advection velocities but was found to be unsatisfactory. As a result, a 'semi-automated' procedure was developed. This process is essentially a manual (point-wise) feature tracking procedure into which the template matching technique which formed the basis of automated procedures, was incorporated as a labour saving device. Tests indicated a time saving of 20-40 % on the manual procedure and more rapid computation than with the automated procedure. The feature tracking procedure was applied to three sets of AVHRR images of the South East Atlantic. To assess the precision of the vector computation procedure, two independent vector sets were computed. A comparison of the two sets indicated that the rootmean- square deviation in vector magnitude (speed) was about 6-8 cm/sand in the vector direction, about 31° (12° if very small vectors ≤ 6 cm/s are excluded). The computed vectors compared very well with reported results from conventional methods. The derived vector fields also provide the first really detailed description of surface currents in the sea off South Africa: eg. on the flow field in the southern Benguela Current, the circulation associated with Agulhas Current rings, and advective influences on the transport of fish eggs and larvae from the spawning grounds on the Agulhas Bank to the favoured recruitment area off the West Coast. |
author2 |
Shillington, Frank A |
author_facet |
Shillington, Frank A Agenbag, Johannes Jacobus |
author |
Agenbag, Johannes Jacobus |
author_sort |
Agenbag, Johannes Jacobus |
title |
A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean |
title_short |
A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean |
title_full |
A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean |
title_fullStr |
A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean |
title_full_unstemmed |
A procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the South East Atlantic Ocean |
title_sort |
procedure for the computation of sea surface advection velocities from satellite thermal band imagery, with applications to the south east atlantic ocean |
publisher |
University of Cape Town |
publishDate |
2017 |
url |
http://hdl.handle.net/11427/23143 |
work_keys_str_mv |
AT agenbagjohannesjacobus aprocedureforthecomputationofseasurfaceadvectionvelocitiesfromsatellitethermalbandimagerywithapplicationstothesoutheastatlanticocean AT agenbagjohannesjacobus procedureforthecomputationofseasurfaceadvectionvelocitiesfromsatellitethermalbandimagerywithapplicationstothesoutheastatlanticocean |
_version_ |
1719331441104388096 |