polar2018.pdf

Transcript

1. Modification of Polar Low Development by Sea Ice and Svalbard Orography Denis Sergeev, Ian Renfrew, Thomas Spengler POLAR2018 | Davos, Switzerland | 22 June 2018
2. What are polar lows?
3. Credit: met.no Credit: U. Tokyo Japan Sea Credit: NEOSDAAS Norwegian Sea Credit: NEOSDAAS Barents Sea Credit: NRK.no Credit: A. Carleton Southern Ocean What are polar lows?
4. Where are they observed?
5. Where are they observed? Stoll et al. (2018) Verezemskaya et al. (2017) Bracegirdle and Gray (2008) Michel et al. (2018)
7. Why the maximum near Svalbard?
8. ? Svalbard ➔ Large non-dimensional height ➔ Major obstacle for CAOs ➔ Upstream blocking, flow splitting and lee vorticity stretching
9. Svalbard ➔ Large non-dimensional height ➔ Major obstacle for CAOs ➔ Upstream blocking, flow splitting and lee vorticity stretching ? Sea ice ➔ Concave shape ➔ Convergence of CAOs ➔ Surface heat fluxes ➔ Is polar low activity correlated with sea ice extent in the North Atlantic? ?
10. How can we test that?
11. How can we test that? - UK Met Office Unified Model to rescue!
12. MetUM nested domain
13. MetUM nested domain 2.2 km
14. “But is it good enough?” - you, probably
15. “Yes, it is*” - Sergeev et al. (2017), QJRMS *As compared to aircraft, dropsonde & satellite observations
16. What polar lows to choose?
17. 1http://polarlow.met.no/stars-dat/ Case A 5-6 April 2007 STARS database1
18. 1http://polarlow.met.no/stars-dat/ Case A 5-6 April 2007 Case B 30-31 January 2008 STARS database1
19. How do these 2 polar lows evolve?
20. Sea ice edge SLP contours (every 2 hPa)
21. Sea ice edge SLP contours (every 2 hPa) Cyclonic vorticity ζ (10-4 s-1) at 950 hPa
22. Sea ice edge SLP contours (every 2 hPa) Cyclonic vorticity ζ (10-4 s-1) at 950 hPa PV at 500 hPa (every 2 PVU)
23. CASE A CASE B First appearance Intensification Mature stage Intensification Mature stage First appearance
24. CASE A CASE B Intensification Mature stage Intensification Mature stage First appearance L Polar low
25. CASE A CASE B Mature stage Intensification Mature stage First appearance L Polar low L Polar low
26. CASE A CASE B Intensification Mature stage First appearance L Polar low L Polar low L Polar low
27. CASE A CASE B Intensification Mature stage L Polar low L Polar low L Polar low L Polar low
28. CASE A CASE B Mature stage L Polar low L Polar low L Polar low L Polar low Polar low
29. CASE A CASE B L Polar low L Polar low L Polar low L Polar low Polar low Polar low
30. Similarities ➔ Western periphery of a synoptic-scale cyclone ➔ Both originate near Svalbard and join vorticity clusters downstream ➔ Develop due to a mix of baroclinic and barotropic instability How CASE A is different to CASE B? ➔ Upper-level PV anomaly >4 PVU ➔ Rossby penetration depth ~4.8 km ➔ Resembles type B cyclogenesis of Petterssen and Smebye (1971)
31. Let’s remove Svalbard!
32. Let’s remove Svalbard!
33. CTRL What we actually did: ➔ Set elevation to 0 ➔ Set land-sea mask to “sea” ➔ Replace with sea ice
34. NOSVA What we actually did: ➔ Set elevation to 0 ➔ Set land-sea mask to “sea” ➔ Replace with sea ice
35. CTRL CTRL NOSVA NOSVA CASE A CASE B Only 1 time step (mature stage)
36. CTRL CTRL NOSVA NOSVA CASE A CASE B Sea ice edge SLP contours (every 2 hPa) Cyclonic vorticity ζ (10-4 s-1) at 950 hPa Only 1 time step (mature stage)
37. NOSVA NOSVA CTRL CTRL CASE A CASE B When Svalbard is removed…
38. NOSVA NOSVA CTRL CTRL CASE A CASE B When Svalbard is removed… ➔ PLs appear still appear ➔ They grow on the Barents Sea vorticity filaments
39. NOSVA NOSVA CTRL CTRL CASE A CASE B When Svalbard is removed… ➔ PLs appear still appear ➔ They grow on the Barents Sea vorticity filaments The vorticity “tail” is detached = no reinforcement from Svalbard CASE A PL is “steered” by the PV anomaly
40. What about sea ice?
41. CTRL ICE76N experiment ➔ Extend sea ice to the south (76°N) ➔ Make the edge zonally straight ➔ (Svalbard is intact)
42. CTRL ICE76N experiment ➔ Extend sea ice to the south (76°N) ➔ Make the edge zonally straight ➔ (Svalbard is intact) ICE76N
43. ICE76N ICE76N CTRL CTRL CASE A CASE B With changed sea ice…
44. ICE76N ICE76N CTRL CTRL CASE A CASE B With changed sea ice… ➔ PLs are weaker ➔ Due to lack of surface heat flux Small-scale waves without a dominant centre
45. ICE76N ICE76N CTRL CTRL CASE A CASE B New layers shown: ➔ Sensible heat flux (positive upward)
46. CTRL CTRL CASE A CASE B Average within a radius for every point along the track
47. CTRL CTRL CASE A CASE B Landfall Landfall Formation Formation Average within a radius for every point along the track Sensible heat flux
48. CTRL CTRL CASE A CASE B Sensible heat flux ICE76N ICE76N
49. CASE A CASE B Sensible heat flux Relative vorticity ➔ The lack of SHF at the start hinders the PL growth ➔ When PL moves to the ice-free surface, there is a big spike in heat flux ICE76N CTRL ICE76N CTRL ICE76N CTRL ICE76N CTRL Ice edge Ice edge
50. CASE A CASE B Sensible heat flux Relative vorticity ➔ The lack of SHF at the start hinders the PL growth ➔ When PL moves to the ice-free surface, there is a big spike in heat flux ➔ However this “boost” of heat fluxes is not enough for the PL to reach CTRL instensity* *Can be an artefact of tracking method (choosing what vorticity cluster to track) ICE76N CTRL CTRL ICE76N CTRL ICE76N ICE76N Ice edge Ice edge CTRL
51. Summary* ➔ Many polar lows develop due to merging of vorticity filaments ➔ The primary sources are convectively-driven convergence lines, while Svalbard provides a secondary source of vorticity ➔ Svalbard deflects polar low tracks, but upper-level PV anomaly can counteract this effect ➔ For our PLs convection is important, so change in sea ice and SST leads to greater changes in PL intensity ➔ Future studies are needed to examine different synoptic situations or sea ice configuration *Submitted to Monthly Weather Review Code is available at github.com/dennissergeev/mplosi
52. Thank you! Credit: Met Norway speakerdeck.com/dennissergeev/polar2018 dennissergeev.github.io github.com/dennissergeev twitter.com/meteodenny