RETROSPECT - ocean analysis using optimal observation strategies

RETROSPECT has been a collaboration between the Norwegian Meteorological Institute (MET Norway), the Institute of Marine Research (IMR) and the Nansen Environmental and Remote Sensing Center (NERSC). The objectives of RETROSPECT have been (1) to better integrate the operational ocean models that are developed nationally and as part of CMEMS, (2) to assess the impact of existing and planned observation networks, and (3) to improve our drift models. The project has funded two postdocs and one PhD student in addition to senior scientists from all three institutions. A data assimilative coastal ocean model based on ROMS-4DVAR was put into operations in December 2017. This model utilizes TOPAZ for boundary conditions, and with a high resolution ROMS-4DVAR system in place and better integration of the Norwegian modeling systems we accomplished our main task for the development of the operational systems. Daily forecasts based on ROMS-4DVAR is now being produced, and we have also contributed to other research projects by assimilating observations from autonomous platforms (GLIDER project, RCN 269188) and ad-hoc observations collected during field campaigns (Norwegian Clean Seas Association Oil-on-water exercise, June 2018). TOPAZ, which is based on the ocean model HYCOM, will be upgraded in April 2019 with the improved physics for wave-mean flow interactions in a version with twice the spatial resolution. NERSC has also contributed a sensitivity impact study for a new proposed concept of satellite SKIM (Sea surface KInematics Multiscale monitoring), which has been selected by the European Space Agency as its potential 9th Earth Explorer mission. ROMS-4DVAR has been used to produce reanalyses for selected periods and regions. One reanalysis covers Lofoten-Vesterålen for an extended period in the mid eighties, during which the IMR made extensive insitu measurements to investigate the fish stocks here. This reanalysis explores the potential impact of using a dense network of profiling instruments, for example such as can be obtained using gliders, and the reanalysis has also been used to assess the benefit of using data assimilative models to study the physical-biological coupling in the upper ocean. The other reanalysis covers Skagerrak and Kattegat in 2017. We had access to HF radar data in this period, and we compared the impact of HF radar surface currents, insitu hydrography and satellite SST on the transport in the Norwegian Coastal Current. We could conclude that HF radar data has a high impact, comparable or larger than the impact of SST. The impact per data point was higher for insitu data, however, but since there are relatively few such observations the total impact was limited. These results support previous findings that a combination of remotely sensed data and insitu data is ideal, and are useful in the planning of a national coastal observation infrastructure. We have published several observation based papers on the upper ocean dynamics, providing more detail about the vertical structure of the near surface layer. We had a dedicated cruise in April 2016 with an ambitious measurement program collecting data on oceanic turbulence, currents, hydrography, airsea fluxes, surface waves and cod egg concentrations. Autonmous instruments were used, and the overall picture is of a complex physical-biological system with large temporal and spatial variability. RETROSPECT researchers also participated an experiment in October 2017, in which drifters were deployed in the coverage area of the Fruholmen HF radar. We have collaborated with leading scientists from abroad, and the project funded to visits abroad for the PhD student and one of the postdocs. Two workshops were arranged gathering all the participants in the project. RETROSPECT researchers have given scientific and popular science presentations both in Norway and abroad. In addition, we co-chaired a session at the 2018 AGU Ocean Sciences Meeting in Portland that was very popular.

I RETROSPECT har vi forbedret de operasjonelle havmodellene. Dette arbeidet har vesentlig styrket tjenesten vi leverer og den direkte samfunnsnytten av prosjektet har vært betydelig. Vi har også gjort viktige framskritt når det gjelder driftsmodellene. Observasjoner av en lang rekke fysiske og biologiske parametere ble samlet inn gjennom to tokt. Disse observasjonene har gitt oss en bedre forståelse av dynamikken i øvre del av havet og de fysisk-biologiske koblingene vi finner her. Vi har også undersøkt nytteverdien av ulike typer observasjoner for dataassimilerende havmodeller, noe som er nyttig for planleggingen av framtidige observasjonsinfrastrukturer. Metodene som er utviklet innenfor prosjektet nå er integrert med de operasjonelle tjenestene. Dette sikrer at resultatene fra RETROSPECT vil ha nytteverdi i overskuelig framtid.

RETROSPECT focuses on the optimal use of ocean observations in numerical ocean models. These models provide data for decision support (e.g. oil spill mitigation and offshore operations), offshore construction design critera and studies of acute and long-term exposure to pollutants in the marine environment. Rapid technological developments open up exciting new possibilities for ocean observations. It is therefore necessary to have tools at hand to devise observation sampling strategies, in particular when the observations are intended for use in data assimilative ocean models. RETROSPECT focuses on northern areas and two field campaigns will provide detailed observations for studying transport and mixing processes at high latitudes. A fully integrated system for devising observation sampling strategies and production of high quality ocean analyses using the field data is proposed. Observation impacts and the role of improved physical forcing for biophysical modeling systems will also be quantified. RETROSPECT brings together the major national oceanographic research institutes, and a strong research group provides expertise on ocean data assimilation, upper ocean dynamics, and biophysical modeling. In addition to the scientific results, project outcomes are new datastreams that will be made publicly available to our users of numerical ocean model data in government and industry.