Fishing from space : mackerel fishing in Icelandic waters and correlation with satellite variables
Author
Summary, in English
As concentrations of measured CO2 in the atmosphere reach a record high it is important to attempt all possible efforts to reduce the emissions of greenhouse gases (GHG) in all aspects of industry. The fishing sector contributes 15% of total GHG emission in Iceland, with the majority originating from fishing vessels using fossil fuel. The relationship between catching locations of Atlantic Mackerel (Scomber scombrus) in Icelandic waters and satellite remote sensing variables was explored. The aim was to provide information for possible fish¬eries forecasting, which could facilitate reduced energy consumption in Icelandic fishing vessels. The hypothesis was that satellite variables were a valuable source of information for determining viable fishing grounds in Icelandic waters.
The variables explored were sea surface temperature (SST), chlorophyll (CHL), photosyn-thetically available radiation (PAR), water leaving radiance (L443) and down welling diffusion attenuation coefficient (kd490). The spatial resolution was about 4.6 km and temporal resolution 1 day. Effects of decreased spatial and temporal resolution were also explored.
Binomial generalized additive models were created to identify the possible relationship with fishing locations represented as absence or presence of mackerel catches. Seven day PAR was the strongest single variable, explaining 47% of deviance, with the spatial variables latitude and longitude incorporated. The most successful multiple variable models included one or seven day averages of PAR and SST and seven day averages of L443, explaining 48% of deviance. Decreasing temporal resolution to 7 days improves the predictive ability of all variables. Decreasing spatial resolution to 3*3 cells does not decrease or increase the predictability to any extent.
In order to estimate the usefulness of global data sets in local situations, a correlation of observed and remotely sensed CHL in Icelandic waters was estimated. Results on a minor sample size revealed a strong significant correlation, suggesting that global datasets were useful in local situations around Iceland.
The satellite variables explored significantly contribute to a model explaining the absence and presences locations for mackerel fishing in Icelandic waters. Mackerel catches were most successful in a temperature range of 7.5°-13°C where there were high amounts of incoming visible solar radiation and intermediate concentration of phytoplankton. Clear waters due to little absorption as well as turbulent water with high scattering also had effects. This suggested that mackerel caught in Icelandic waters was more dependent on visual foraging than previously considered.
The variables explored were sea surface temperature (SST), chlorophyll (CHL), photosyn-thetically available radiation (PAR), water leaving radiance (L443) and down welling diffusion attenuation coefficient (kd490). The spatial resolution was about 4.6 km and temporal resolution 1 day. Effects of decreased spatial and temporal resolution were also explored.
Binomial generalized additive models were created to identify the possible relationship with fishing locations represented as absence or presence of mackerel catches. Seven day PAR was the strongest single variable, explaining 47% of deviance, with the spatial variables latitude and longitude incorporated. The most successful multiple variable models included one or seven day averages of PAR and SST and seven day averages of L443, explaining 48% of deviance. Decreasing temporal resolution to 7 days improves the predictive ability of all variables. Decreasing spatial resolution to 3*3 cells does not decrease or increase the predictability to any extent.
In order to estimate the usefulness of global data sets in local situations, a correlation of observed and remotely sensed CHL in Icelandic waters was estimated. Results on a minor sample size revealed a strong significant correlation, suggesting that global datasets were useful in local situations around Iceland.
The satellite variables explored significantly contribute to a model explaining the absence and presences locations for mackerel fishing in Icelandic waters. Mackerel catches were most successful in a temperature range of 7.5°-13°C where there were high amounts of incoming visible solar radiation and intermediate concentration of phytoplankton. Clear waters due to little absorption as well as turbulent water with high scattering also had effects. This suggested that mackerel caught in Icelandic waters was more dependent on visual foraging than previously considered.
Department/s
Publishing year
2013
Language
English
Publication/Series
Student thesis series INES
Full text
- Available as PDF - 15 MB
- Download statistics
Document type
Student publication for Master's degree (two years)
Topic
- Earth and Environmental Sciences
Keywords
- chlorophyll
- downwelling diffuse attenuation coefficient
- photosynthetically available radiation
- water leaving radiance
- physical geography and ecosystem analysis
- sea surface temperature
- Atlantic Mackerel
- Scomber scombrus
- Icelandic waters
- satellite variables
- remote sensing
Funder
- AVS R&D Fund of Ministry of Fisheries in Iceland
- Landsvirkjun Energy Research Fund
Report number
279
Supervisor
- Jonathan Seaquist