Airborne remote sensing for velocity field determination
SR(00)13
August 2001
Executive summary
- To understand and manage the effects of pollution in estuarine and near-shore coastal waters requires an understanding of the often complex circulation processes of such systems. At present, this knowledge is derived from a combination of limited in situ field methods and the output from mathematical models.
- To understand the complexity of the velocity fields operating in the coastal zone, comprehensive spatial and temporal datasets are required. Maps of variables such as temperature and chlorophyll-a, derived from airborne remotely sensed data, have the potential to meet the spatial and temporal resolutions required with which to validate the output from circulation models and to overcome the limitations of in situ sampling.
- A number of methods have been developed by which the velocity fields have been determined from mesoscale remotely sensed data (e.g. AVHRR imagery). To date these have not been tested on airborne high spatial resolution data for near-shore waters. One such method is Maximum Cross Correlation (MCC), a pattern matching technique which assigns vectors to indicate the maximum displacement in patterns of water quality between successive images.
- The aim of this study was to investigate the utility of repeat-pass airborne remotely sensed data and Maximum Cross Correlation (MCC) for deriving accurate surface velocity maps for an estuarine environment.
- Repeat-pass Airborne Thematic Mapper data were acquired at approximately 10 minute intervals on two occasions over Kirkcudbright Bay in Galloway, a small estuary showing strong tidal movements. On one of those occasions, water movements were also monitored using surface drogues which recorded maximum velocities of approximately 0.4 m s-1.
- Sequences of whole-bay and subsections of remotely sensed images illustrated that greatly improved levels of hydrodynamic detail could be qualitatively inferred from them through the movement and evolution of eddy and frontal features during the tidal cycle. Although thermal data appeared more useful, both thermal and chlorophyll maps showed qualitatively similar motions.
- Synoptic velocity fields derived from MCC concurred with the south and eastward displacement seen through the qualitative interpretation of images. Velocities calculated were all below 0.6 m s-1 which concur well with the in situ drogue measurements. Sources of error are thought to relate to inaccurate geopositioning of the images with respect to one another.
- Synoptic velocity fields derived for images processed to indices of Sea Surface Temperature and Chlorophyll-a were similar but subtly different. Differences are to be expected as chlorophyll-a images are derived from a greater part of the water column (top few metres) than the temperature images (top few micrometres). Chlorophyll-a images may therefore include sub-surface water velocity effects.
- An evaluation of increasing the time interval between images at six intervals ranging from 5 to 32 minutes showed consistent increases estimated displacement for a single test cell. However, there was also an increase in the number of cells that produced discordant velocity vectors which needed removal by filtering.
- The results indicate the significant potential of airborne remote sensing for providing improved detail on hydrodynamic processes compared to in situ sampling measurements.
- Further avenues for research were indicated. These include improvements to the MCC technique to remove the chance of obtaining spurious vectors, the use of neural networks as an alternative to MCC methods, the integration of velocity field data with hydrodynamic models to provide extended temporal coverage, and techniques for the collection of improved ground validation data.
- Academic presentations and publications arising out of this research are presented in Appendix 2.. These include a paper in an international journal and the proceedings of an international conference.
Copies of this report are available from the Foundation, price £35.00, less 20% to FWR members