|
Játtað í: 2021 Granskingarøki: Náttúra og náttúrutilfeingi Verkætlanarslag: Post.doc. Verkætlanarheiti: Forritan av fjarða myndilium FarCoast og háavrikandi teldumegi Játtanarnummar: 0464 Verkætlanarleiðari: Sissal Erenbjerg Stovnur/virki: Firum (Fiskaaling) Aðrir luttakarar: Gunnvør á Norði, Jon Albretsen, Bárður Niclasen, Eigil Kaas, Karin Margretha H. Larsen, Bogi Hansen og Tróndur T. Johannsen Verkætlanarskeið: Planned: 01.10.2021 - 01.10.2023 Actual: 01.01.2022-31.08.2026 Samlaður kostnaður: kr. 1.686.749 Stuðul úr Granskingargrunninum: kr. 940.000 Verkætlanarlýsing: Original This PostDoc project will focus on creating a HPC center at Fiskaaling (FA), gaining the best possible performance of existing and new super-computers. This HPC environment will be implemented in close collaboration with the University of the Faroe Islands (UoFO), the Institute of Marine Research in Norway (IMR) and the University of Copenhagen (UCPH). The establishment of a HPC environment will be beneficial for the entire Faroese research community, as we wish to share computational resources between our institutes and continue to benefit from these cross disciplines. The contribution of the Post Doc will be optimizing the FarCoast model. FarCoast is adapted from NorKyst800, a model covering the Norwegian coast using 800m x 800m resolution (Asplin et al. 2020 and Albretsen et al. 2011), which is based on the Regional Ocean Model System (ROMS). FarCoast is a high-resolution 3D ocean model for the Faroese coasts and fjords. FarCoast has a triply nested grid (800m –> 160m –> 32m), increasing resolution down to 32m in the fjords. The FarCoast800 model setup is validated by Erenbjerg et al. (2020) who concludes that the ROMS is an excellent choice for driving a high-resolution nested model of the Faroe Shelf, since FarCoast800 provides circulation in accordance with observations on the shelf (Erenbjerg et al. 2020). Working results suggest that the FarCoast32 with 32x32 m grid resolution applied in Sundalagið North can realistically simulate current velocities and exchanges with shelf water (correlation coefficient: r > 0.8 (In prep Erenbjerg et al.)). The importance of having such a hydrodynamic model running in the Faroe Islands is imperative. It can be widely applied as a driver for other near-field models, investigating the Faroese enviromnent and state of nature. Fiskaaling has invested considerable work effort in an ongoing endeavor to understand salmon lice dispersion and infection pressure (á Norði et al. 2015, 2016, Kragesteen et al 2018, 2019, 2021). The ability to combine FarCoast with an individually based model for salmon lice, will increase the realistic simulation of lice dispersion as dispersion patterns are primarily determied by hydrography and current velocities (Sandvik et al. 2016, Asplin et al. 2014, Johnsen et al. 2016, Myksvoll et al. 2020). Final The project has established and strengthened a sustainable computational and ocean-modelling environment at Firum for high-resolution modelling of Faroese shelf, coastal and fjord waters. The computational infrastructure was developed during the project and is now based on a Proxmox environment at Firum, providing a flexible platform for numerical modelling and related computational activities. A major scientific result of the project is the development and validation of a 2 km regional FarCoast hydrodynamic model and an associated 20-year hindcast archive for Faroese waters. The original project aimed to establish a 10-year hindcast archive, and the resulting 20-year archive therefore represents a substantial extension of the planned model dataset. In parallel, a 130 m high-resolution FarCoast model has been developed for more detailed representation of coastal and fjord processes. Validation of this model is currently underway. The first validation results and technical documentation are planned for publication in a Firum Technical Report in 2026, followed by a separate peer-reviewed scientific publication. An important improvement compared with the original Data Management Plan has been the establishment of a local THREDDS data server at Firum. This provides structured and efficient access to the large model archive and facilitates the reuse of model results across research projects. The modelling infrastructure, model output and expertise developed through the project are already being applied in several research projects at Firum, including Adepoplan, FjordProcess, GloBecc, OCCAM and BlueCea. Applications include physical oceanography, particle tracking, sea-lice dispersion, seaweed transport and other marine and aquaculture-related modelling. An important long-term outcome is the establishment of a broader and more permanent modelling environment at Firum. Three staff members currently work directly with numerical modelling, with complementary expertise in hydrodynamic ocean modelling, particle tracking and broader modelling software development. International collaboration has also been strengthened. The close collaboration with the Institute of Marine Research, Norway, has continued through regular scientific exchange, including a research visit by the Project Manager in 2023. New collaborations have also developed with modelling environments in Iceland and Scotland. In Iceland, ROMS expertise developed through the project is now being used to support researchers in establishing and developing their own local ROMS modelling system. Three bachelor projects have also been supervised within the framework of the project, contributing to education and local capacity building in numerical modelling and oceanography. Støða: Liðug Avrik: See PURE Storage and access rights to collected data The project has generated a large numerical archive of hydrodynamic model output, including the 20-year hindcast from the 2 km regional model. Model data are stored at Firum and organised for access through Firum’s THREDDS Data Server, providing structured access to model datasets and facilitating their reuse in subsequent research and modelling applications. See THREDDS Data Server In accordance with the project’s Data Management Plan, the project consortium retains a two-year publication privilege for newly generated model data. Following scientific publication and the agreedpublication period, relevant model products will be made available for wider scientific use at the TREDDS server. << Back |
| Administration English |