2021
DOI: 10.1002/qj.3971
|View full text |Cite
|
Sign up to set email alerts
|

Central Arctic weather forecasting: Confronting theECMWF IFSwith observations from the Arctic Ocean 2018 expedition

Abstract: Forecasts with the European Centre for Medium‐Range Weather Forecasts' numerical weather prediction model are evaluated using an extensive set of observations from the Arctic Ocean 2018 expedition on the Swedish icebreaker Oden. The atmospheric model (Cy45r1) is similar to that used for the ERA5 reanalysis (Cy41r2). The evaluation covers 1 month, with the icebreaker moored to drifting sea ice near the North Pole; a total of 125 forecasts issued four times per day were used. Standard surface observations and 6‐… Show more

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

3
29
0

Year Published

2022
2022
2024
2024

Publication Types

Select...
7
1

Relationship

3
5

Authors

Journals

citations
Cited by 30 publications
(32 citation statements)
references
References 49 publications
3
29
0
Order By: Relevance
“…For points above the boundary layer no further distinction was made between land, ocean, or ice. The Arctic boundary layer is difficult to accurately represent in models (Birch et al., 2012 ; Graham et al., 2019 ; Sotiropoulou et al., 2016 ; Tjernström et al., 2021 ; Young et al., 2021 ). However, varying the ECMWF boundary layer pressure by ±15 hPa or assuming a constant value (965 hPa, which corresponds to the mean boundary layer pressure along all trajectory data points started from the lowest model layer) did not significantly change the results or the patterns observed (see Figure S2 in Supporting Information S1 ).…”
Section: Methodsmentioning
confidence: 99%
“…For points above the boundary layer no further distinction was made between land, ocean, or ice. The Arctic boundary layer is difficult to accurately represent in models (Birch et al., 2012 ; Graham et al., 2019 ; Sotiropoulou et al., 2016 ; Tjernström et al., 2021 ; Young et al., 2021 ). However, varying the ECMWF boundary layer pressure by ±15 hPa or assuming a constant value (965 hPa, which corresponds to the mean boundary layer pressure along all trajectory data points started from the lowest model layer) did not significantly change the results or the patterns observed (see Figure S2 in Supporting Information S1 ).…”
Section: Methodsmentioning
confidence: 99%
“…Smallscale physical atmospheric processes are often poorly represented in models at all spatial and temporal scales. For example, warm biases in models at low levels have been associated with errors in Arctic surface sensible heat fluxes (Tjernstro ¨m et al, 2021). Additionally, the long-standing model challenge of representing Arctic clouds can lead to biases in modeled radiative balance and feedbacks (Karlsson and Svensson 2013;Urrego-Blanco et al, 2019;Kretzschmar et al, 2020).…”
Section: Introductionmentioning
confidence: 99%
“…The atmospheric model is compiled with either default values for roughness length and skin conductivity or with adapted surface conditions to more closely follow observations. To match the observed friction velocity (similar to Sotiropoulou et al [2018]) the roughness length z 0 is increased from 10 −3 to 0.03 m. To enforce stronger coupling with the surface, and to correct a known warm bias in the model under melting conditions (Tjernström et al, 2021), the skin conductivity Λ is increased to basically represent instantaneous coupling, from 58 to 10 10 Wm −2 K −1 (see also Section 3.1). The alternative skin conductivity value for the sea ice is equivalent to the default value over the ocean.…”
Section: Methodsmentioning
confidence: 99%
“…For example, over the ocean a large skin conductivity represents negligible heat storage in the skin layer while over sea ice a much lower value was set originally. However, over melting sea ice in Arctic summer conditions the surface temperature is constrained by phase change, and a lower skin conductivity then leads to too high near‐surface temperatures in IFS (Tjernström et al., 2021).…”
Section: Model Description and Data Sourcesmentioning
confidence: 99%