Exploiting international synergies with PREFACE to enhance understanding of the relationship between regional and global model errors

The core objective of PREFACE was to improve predictions for the tropical upwelling system of the Benguela (southern Africa) on seasonal to decadal time scales. The predictions with numerical ocean-atmosphere global circulation models are dominated by drift towards a model climatology that differs substantially from the observed one. Uni Research's participation within PREFACE is dedicated to understanding the mechanisms and causes of this drift. The NFR synergies grant had the objective of supporting this activity via international collaborations. An important source of model error was identified in the spatio-temporal pattern of the winds that drive the regional ocean circulation. The initial part of the project thus focussed on empirically diagnosing model biases, attempting to formulate hypothese as to their origin, and initiating code development work that could improve the model. This activity has progressed well, but it also led to understanding that the problem is deeper and more fundamental than initially thought. First, it is not known which aspects of the wind forcing actually affect ocean currents and water-mass properties of the Benguela. Second, it is not known what mechanisms are controlling the coastal surface winds. Third, there are virtually no in-situ observations of the atmospheric state in the Benguela that can guide model improvements and help understanding the likely effects of climate change. The project therefore evolved towards addressing these knowledge gaps. In year 2, in partnership with the University of Cape Town efforts were made towards a theory of the response of the coastal ocean to wind forcing, and of the nature of that forcing. It was realised that there are no data that can provide support or falsification of that theory. This led to an intensive observations effort that has involved partnerships with the FAO, the Norwegian HI, and the Namibian Ministry of Fisheries, in addition to the exising Norwegian, European, and South African partners. Over November and December 2017, TT performed 53 atmospheric sounding off the Namibian coast from the R/V Dr.Fridtjof Nansen. Additionally, the grant was used to establish links with the international community via the CLIVAR RF group on Eastern Boundary Upwelling System. TT is now leading the data-model comparison effort as co-chair of that group. In a team of top international scientists we will gather all available data sources and diagnose the quality of reanalysis and models in the Benguela and other EBUSs, bringing together new observations, new theory, and improved models.

Et hovedmål for EU-prosjektet PREFACE, der Thomas Toniazzo leder modelleringsarbeidet i Kjernetema 3, er å forbedre simuleringen av klimaet i det tropiske Atlanterhavet ved å adressere systematiske feil og biaser i klimamodellene. Samtidig vil EU-prosjekt et SPECS adressere globale prediksjoner på sesong til tiårs-tidsskala basert på numeriske sirkulasjonsmodeller, inkludert deres systematiske biaser. ESPERGE foreslår en ytterligere styrking av forskningsaktiviteten innenfor PREFACE gjennom europeisk samar beid med SPECS. I ESPERGE vil dette hovedsakelig bli oppnådd gjennom støtte til møter mellom partnerene i de to prosjektene. Målet med dette er å bidra til økt utveksling av vitenskapelige resultater samt å gjøre modellsimuleringer fra de to prosjektene gjensidig tilgjengelig. ESPERGE vil også bidra til kunnskapsutveksling om vitenskapelige og tekniske aspekter ved formulering, konfigurering og validering av modelleksperimenter. Et resultat av en slik kunnskapsutveksling vil være gjennomføringen av eks tra dedikerte sensitivitetseksperimenter med sikte på adressere modellbiaser i simuleringer av nedbør over nord-Afrika. I tillegg vil modellevalueringen og utviklingen av NorESM bli ytterligere styrket gjennom økt kunnskapsutveksling med CAM utviklerene ved NCAR og ETOS arbeidsgruppen innenfor CLIVAR. Dette vil også bidra til en økt profilering av modellutviklingsarbeidet av sesong og dekade prediksjon ved Uni Research som er planlagt innenfor PREFACE.