TITLE
Assessment of the application of multibeam echosounder technologies to UK marine
habitat classification systems.
SUPERVISOR(S)
Dr. Gay Mitchelson-Jacob, Dr. Ronan Roche, Dr. Jim Bennell
COLLABORATION/CASE PARTNERS
Is it expected that this project would lead to collaborations with marine ecological survey
companies such as Marine Ecological Solutions Ltd., based in Menai Bridge.
BACKGROUND/CONTEXT OF PROJECT
Multibeam echosounder (MBES) systems are widely recognised as an effective tool for
seafloor mapping, but the quantitative characterisation of MBES data and its application to
ecologically relevant classification systems is less developed. Whilst the data availability and
spatial resolution of MBES systems are advancing, the computing power to analyse large
multivariate physical and ecological survey datasets is also increasingly available.
The acoustic backscatter data captured by MBES has application in mapping of surficial
sediments, and methods have been developed to objectively classify backscatter information
into acoustically similar regions (Brown et al., 2011). Ground-truthing of results has shown
variable ability of such systems to correctly identify sediment characteristics and indicates
that improvements can be made in classification techniques.
Physical gradients in particle size are correlated with biological community composition, and
an emerging body of research is investigating the ability of acoustic backscatter data to
identify marine biological assemblages. Considerable variability has been shown in the
strength of correlations observed between acoustic signal-based classification and ground
truth video data. However, high levels of correlation are observed for specific benthic
habitats such as Sabellaria reef, seagrass meadows and mussel beds (e.g. van Rein et al.,
2011). Questions remain as to the degree of detail with which acoustic backscatter data can
be applied to ecological mapping and the extent to which its inclusion in multi-method
classification techniques is useful.
At present, many elements of assigning biotopes to an area based on ecological survey results
are heavily based upon expert judgment. This has the disadvantage of being a subjective
methodology, such that the final biotope choice may be difficult to defend. There is an
opportunity for developing statistical tools to support the decision making process and to
incorporate additional types of data such as acoustic backscatter for achieving more robust
objective outcomes in the biotoping process.
AIMS AND OBJECTIVES
Aims:
1. To investigate the ability of MBES data to contribute to biotope mapping, by
comparing classification based on acoustic backscatter data with ground-truthing
methods.
2. To develop novel techniques to support the decision-making process in the marine
biotoping process.
Objectives:
1. To carry out ship-based surveys of marine habitat incorporating MBES data
acquisition, video (sledge or drop-down) and sediment grab-sampling techniques.
2. To examine methods for refining systems of marine habitat classification based on
acoustic backscatter data.
3. To rigorously compare results obtained with video and sedimentological data.
4. To analyze the existing database of acoustic backscatter data, in order to test the
application of classification methods on a range of sedimentary environments.
5. To develop techniques for improving the marine biotoping classification process, by
testing the potential contribution of biological abundance measurements, and acoustic
backscatter data.
MATERIALS/METHODS/TRAINING ELEMENT
Materials and Methods:
This project would expand the application of the MBES systems available on board the R/V
Prince Madog and the Macoma. Considerable amounts of acoustic backscatter data have been
collected during cruises, but the majority has not yet been analysed.
In addition to working with this database of existing data, it is expected that this project
would focus on collection of new data from a suitable site as a basis for comparative analysis.
Training methods would include:
1. Training in ship-based collection of sediment grab samples, video surveys, and multibeam data acquisition.
2. Training in MBES data processing methods using the PDS 2000 software.
3. Training in lab-based sedimentological analysis, video survey analysis and GIS
mapping techniques.
4. Development of a familiarity with the UK MNCR biotoping classification system, as
well as the EU EUNIS classification system.
COSTS
As this project is related to SEACAMS R & D project(s), most of the project costs will be
met through SEACAMS.
Prince Madog cruise time. It is expected that the PhD would entail participation in 2
SEACAMS cruises of approximately five days duration aboard the R/V Prince Madog.
Macoma boat time. This project may additionally require 2 weeks of SEACAMS Macoma
time for MBES data collection, and collection of sediment and drop-down video data for
ground truthing.
Consumables associated with in situ sampling will be met through SEACAMS.
There are software requirements to enable processing of MBES acoustic backscatter data.
Some classification may be done using the PDS 2000 and Fledermaus Geocoder software
which SOS has, however software such as QTC Swathview would considerably expand the
ability to process and classify acoustic backscatter data. A quote has been received for QTC
Swathview which is ~£12,000.
References:
Brown, Craig J.; Todd, Brian J.; Kostylev, Vladimir E. (2011). Image-based classification of
multibeam sonar backscatter data for ojbective surficial sediment mapping of Georges Bank, Canada.
CONTINENTAL SHELF RESEARCH Volume: 31 Issue: 2 Special Issue: SI Supplement: S
Pages: S110-S119 DOI: 10.1016/j.csr.2010.02.
van Rein, H.; Brown, C. J.; Quinn, R. (2011) An evaluation of acoustic seabed classification
techniques for marine biotope monitoring over broad-scales (>1 km2) and meso-sclaes (10 m2 -1 km2).
ESTUARINE COASTAL AND SHELF SCIENCE Volume: 93 Issue: 4 Pages: 336-349 DOI:
10.1016/j.ecss.2011.04.011