1. Convective initiation ahead of squall lines Robert Fovell UCLA Atmospheric & Oceanic Sciences rfovell@ucla.edu (Fovell, Mullendore and Kim 2006, MWR)

2. Radar image of a squall line

3. Vertical cross-section

6. A typical multicellular squall line

7. Vertical cross-section “discrete convective initiation”

8. Vertical cross-section

9. “discrete propagation” X

10. 7 May 1995, early evening

11. 0509Z - Hastings, NE radar 8 July 2003 gust front

12. 0539Z - Hastings, NE radar new cells ~ 18 km ahead

13. 0549Z - Hastings, NE radar

14. 0609Z - Hastings, NE radar

15. Animation of Hastings radar

16. X = Hays, KS 05 June 2004

17. X = Hays, KS 05 June 2004

18. 21 June 2003, W Oklahoma ~ after midnight

19. 2245Z (545 PM CDT)

21. 2245Z (545 PM CDT) Ft. Worth

22. 00Z Fort Worth hodograph

23. Rolls in an ARPS simulation

24. How do afternoon roll clouds influence nocturnal convection? By organizing the moisture field; effect survives rolls themselves

25. MM5 simulation 4 km horizontal resolution; 250x330 pts Start 12Z previous day Initial/boundary conditions from Eta model MRF PBL scheme

26. MM5 model animation 3 hour animation (01-04Z) Colored field is 2 m water vapor Vertically integrated condensate contoured 10 m wind vectors

28. MM5 moisture bands Remains of convective rolls present in model on previous afternoon Rolls are spurious –reflect deficiency of PBL scheme –~40 km wavelength >> theoretical value –actual roll clouds ~ theoretical value Rolls are fortuitous –suggest orientation for the new cell lines

29. “Action at a distance” mechanism Trapped internal gravity waves

30. An ARPS simulation 2D & 3D models Horizontally homogeneous initial conditions ∆x = 1 km, ∆z ≥ 40 m Warm rain processes Starts late afternoon

31. Vertical velocity (colored) ~ sunrise main updraft cold pool

32. Vertical velocity (colored) ~ sunrise gravity waves 20 m/s

33. Vertical velocity (colored) ~ sunrise Trapping or ducting below 8-9 km

34. Vertical velocity (colored) ~ sunrise

35. Gravity wave ducting z x Scorer parameter

36. Closer look at Scorer parameter In mountain wave derivation, we had Difference: mountain waves presumed steady, therefore  = 0 and c =  /k = 0. Also, N*2 is BV frequency modified for moisture.

37. Ducting: sharp decrease of l 2 with height Here c > U Forward anvil as wave duct –decrease in ambient stability anvil: warming below, cooling above; saturated partially opposed by (U - c ) decrease –jet-like wind profile - curvature shear

38. upstream sounding

39. U zz min

40. upstream sounding

41. Trapped waves leading to discrete initiation

42. 6 h ARPS model animation

43. Discrete initiation by gravity waves alone Note forward anvil

44. Discrete initiation by gravity waves alone Gravity waves trapped beneath anvil

45. Discrete initiation by gravity waves alone Wave-relative flow shown (recall c > U)

46. Transient trapping conditions

47. Internal gravity waves alone apparently can’t account for the orientation of the new cell bands Combine gravity waves & moisture bands

48. Hypothesis Plane view, looking from above Moisture bands remaining from earlier roll activity

49. Hypothesis Squall line and its forward anvil

50. Hypothesis Trapped internal gravity waves beneath anvil

51. Hypothesis Moisture bands preferred locations for discrete initiation

52. Hypothesis Expect newest cells farthest away along moisture band

55. Summary A case of discrete initiation has been observed & simulated using variety of models New cell lines may be forming along pre- existing moisture bands left by previous roll activity “Action at a distance” may be provided by internal gravity waves excited by main storm Available observations appear insufficient to confirm or refute this hypothesis