A gridded dataset (SMHI Gridded Climatology - SMHIGridClim) has been produced forthe years 1961 - 2018 over an area covering the Nordic countries on a grid with 2.5 kmhorizontal resolution. The variables considered are the two meter temperature and twometer relative humidity on 1, 3 or 6 hour resolution, varying over the time periodcovered, the daily minimum and maximum temperatures, the daily precipitation and thedaily snow depth. The gridding was done using optimal interpolation with the gridppopen source software from the Norwegian Meteorological Institute.Observations for the analysis are provided by the Swedish, Finish and Norwegianmeteorological institutes, and the ECMWF. The ECA&D observation data set (e.g. usedfor the gridded E-OBS dataset) was considered for inclusion but was left out because ofcomplications with time stamps and accumulation periods varying between countries andperiods. Quality check of the observations was performed using the open source softwareTITAN, also developed at the Norwegian Meteorological Institute.The first guess to the optimal interpolation was given by statistically downscaledforecasts from the UERRA-HARMONIE reanalysis at 11 km horizontal resolution. Thedownscaling was done to fit the output from the operational MEPS NWP system at 2.5km with a daily and yearly variation in the downscaling parameters.The quality of the SMHIGridClim dataset, in terms of annual mean RMSE, was shown tobe similar to that of gridded datasets covering the other Nordic countries; “seNorge”from Norway and the dataset “FMI_ClimGrid” from Finland.

Ett klimatologiskt griddat datasett (SMHI Gridded Climatology - SMHIGridClim) hartagits fram för åren 1961 – 2018. Data täcker de nordiska länderna med en horisontellupplösning av 2,5 km. Variablerna som tagits fram är lufttemperatur och relativluftfuktighet vid 2m höjd med en upplösning av1,3 eller 6 timmar beroende av tidsperiod,samt dygnsupplöst min- och maxtemperatur, nederbörd och snödjup. Datasetet ärframtaget med optimal interpolation av stationsdata genom analysverktyget gridpp, somär en öppet tillgänglig programvara från Norska Meteorologiska Institutet.Observationer till analysen har erhållits från de svenska, norska och finskameteorologiska instituten, samt ECMWF. En ansats gjordes också att användaobservationer från datasetet ECA&D från KNMI, men på grund av svårigheter med atttidsstämplarna för data från olika länder inte överensstämde, uteslöts datasetet uranalysen. Kvalitetskontroll av observationerna gjordes med programvaran TITAN, somäven den finns tillgänglig från och utvecklats av Norska Meteorologiska Institutet.Som en första gissning till interpolationen användes statistiskt nerskalade prognosfält(från 11 km till 2,5 km upplösning) från UERRA-HARMONIE. Nerskalningen gjordesmot fält från den operationella numeriska väderprognosmodellen MEPS. Anpassningengjordes med nedskalningsparametrar som varierar över året och dygnet.Kvalitén hos ”SMHIGridClim med avseende på genomsnittligt RMSE är liknande densom tagits fram för griddade data för andra nordiska länderna med varierandeanalysmetoder; “seNorge” från Norge och “FMI_ClimGrid” från Finland.

Recent accelerated warming over the Arctic coincides with sea ice reduction and shifting patterns of land cover. We use a state-of-the-art regional Earth system model, RCAO-GUESS, which comprises a dynamic vegetation model (LPJ-GUESS), a regional atmosphere model (RCA), and an ocean sea ice model (RCO), to explore the dynamic coupling between vegetation and sea ice during 1989-2011. Our results show that RCAO-GUESS captures recent trends in observed sea ice concentration and extent, with the inclusion of vegetation dynamics resulting in larger, more realistic variations in summer and autumn than the model that does not account for vegetation dynamics. Vegetation feedbacks induce concomitant changes in downwelling longwave radiation, near-surface temperature, mean sea level pressure, and sea ice reductions, suggesting a feedback chain linking vegetation change to sea ice dynamics. This study highlights the importance of including interactive vegetation dynamics in modeling the Arctic climate system, particularly when predicting sea ice dynamics. Plain Language Summary Recent accelerated warming over the Arctic is associated with dramatic changes in the physical environment, among which unprecedented sea ice decline has received particular attention. In this study, we use a regional Earth system model accounting for interactive coupling between the atmosphere, land vegetation, and sea ice dynamics to explore their potential links. Our model simulates observed spatiotemporal patterns of sea ice thickness and extent reasonably well. Furthermore, the results show that feedbacks of warming-driven vegetation changes on the near-surface radiation balance can cause greater variations in sea ice between seasons, which can contribute to an accelerated trend of sea ice reduction. The changes in mean sea level pressure caused by vegetation changes can alter the transport of energy and warm the land, sea, and sea ice surfaces. Downwelling longwave radiation is the dominant factor contributing to the near-surface warming and increased sea ice melting. Our study highlights the importance of adopting fully coupled Earth system models that account for interactive effects of vegetation dynamics on the physical climate system, in particular when analyzing the reduction of sea ice in the Arctic.