KRILLSCAN: An automated open-source software for processing and analysis of echosounder data from the Antarctic krill fishery

Summary

Krillscan is an open-source software developed to automatically process echosounder data from fishing vessels to calculate krill biomass.When scientists tested Krillscan against the traditional manual method of analyzing the data, it gave nearly identical results.The software enables data compression (up to 100 times), near real-time analysis, and remote data transfer, allowing fishing vessels to contribute to krill biomass surveys without requiring specialized scientific personnel on board.

The figure shows the software's graphical user interface for reviewing underwater sound data. The interface displays underwater sound images (echograms) that show where krill swarms are located in the water column. Users can view both the raw underwater sound data and the processed version, and can manually correct any detection errors. This visual checking tool ensures the accuracy of the automated krill measurements before the data is transmitted via satellite from fishing vessels to research centers on land.

Key Findings

Krillscan automatically analyzes underwater sound data (echosounders) to estimate krill biomass

The software underestimated krill biomass by only 6% compared to traditional manual methods, with a coefficient of variation of 0.06

Data can be compressed up to 100 times, enabling efficient transfer via satellite

The software processes data faster than it's collected, allowing near real-time analysis

Enables fishing vessels to conduct acoustic surveys without requiring scientific personnel on board

Abstract

Krillscan software was developed to automatically process echosounder data and achieve an accelerated and transparent analysis of backscatter data that allows calculation of target biomass. Herein, the fishery for Antarctic krill (Euphausia superba, Henceforth Krill) was used as a case study to develop the approach. Implementation of a sustainable management strategy for the krill fishery is complicated by a lack of regularly updated krill abundance data on spatiotemporal scales of the fishery. To increase krill biomass data availability, automatic echosounder data processing and swarm detection software was tested against traditional manual scrutinization with LSSS software and agreed with only minor offsets in estimated nautical area scattering coefficients. In addition to automatic processing and data transfer, Krillscan also has a graphical user interface to supervise automatic krill swarm detection. Echogram size can be compressed up to 100 times and raw data are processed faster than generated, thereby enabling near-real time analysis and data transfer. Compressed data can be transmitted online to allow fishing vessels to conduct surveys without having scientific personnel with special expertise on board.

Published in

Fisheries Management and Ecology

2024

Authors

Menze, S., Macaulay, G. J., Zhang, G., Lowther, A. D., & Krafft, B. A.

Institutions

Institute of Marine Research, BergenAqualyd Limited, WakefieldNorwegian Polar Institute, Tromsø

Methods

AcousticData

Read the full paper

Go to doi.org