Doctoral defence – Warsha Singh

Benedikt Magnússon, June 12, 2015
When: Friday, June 12, 2015 – 10:00
Location details: The Aula

Háskóli ÍslandsOn Friday the 12th of June Warsha Singh will defend her Ph.D. thesis in Ecological Modelling. The thesis is titled: Towards efficient benthic survey design with the use of Autonomous Underwater Vehicles.

The defence will take place in the Aula at the Main Building and starts at 10:00

Opponents are dr. Michael Fogarty, Chief of the ecosystem management program, Northeast fisheries science center, USA and dr. Arthur Trembanis – Associate professor of oceanography and geological sciences, University of Delaware, USA.

The supervisor is dr. Gunnar Stefánsson, Professor at the Faculty of Physical Sciences at the University of Iceland. The doctoral committe also includes Dr. Erla Björk Örnólfsdóttir, Dean of Hólar University College, dr. Jörundur Svavarsson, Professor at the Faculty of Life and Environmental Sciences, University of Iceland and dr. Tómas Phillip Rúnarsson , Professor at the Faculty of Industrial Engineering, Mechanical Engineering and Computer Science , University of Iceland.

The ceremony will be chaired by dr. Hafliði Pétur Gíslason, Professor and the head of Faculty of Physical Sciences at the University of Iceland

Abstract

This research work contributes towards improvement of stock assessment techniques for macrobenthic organisms in Icelandic waters with the use of an autonomous underwater vehicle (AUV) as a survey tool for population assessments, and through considering ways in which designs of such surveys can be made more efficient. The Iceland scallop Chlamys islandica (O.F. Müller) population in West Iceland was used as an instructive example to develop the use of a Gavia AUV for benthic research purposes in Icelandic waters.

A method for quantitative population assessment of Iceland scallops from AUV photos was developed. It was shown that small-scale AUV surveys can be repeated in a feasible manner to gather enough data replicates to estimate variance of key population parameters, such as mean abundance and size distribution. Information on variability can be used to optimize survey designs by calculating the number of samples that would yield acceptable survey precision. A modest comparison of scallop size distributions obtained from a AUV and classical dredge survey highlighted the bias and size-selectivity of the dredge survey. Optimized sampling strategies for length distributions were also evaluated and emphasis was placed on the importance of detecting peaks (possible year classes) in the distribution with certainty. A generic approach was presented that incorporates sampling costs to identify the optimum number of samples and sample sizes to achieve this. Further, habitat classification techniques for automated detection of scallop beds from AUV images were evaluated. The mechanism developed can potentially be used to automatically classify a large set of seafloor photos to detect scallop habitats.

The thesis is methodological and does not necessarily draw any biological conclusions from the study. The analytical techniques developed are generic and can be applied to most projects of this nature. To the best of our knowledge, this is the first time an AUV has been successfully used for quantitative fisheries stock assessment purposes in Icelandic waters.