1. Non-baroclinic Inland Rejuvenation of Tropical Cyclones Kerry Emanuel Massachusetts Institute of Technology with special thanks to Jeff Callaghan and Peter Otto, Bureau of Meteorology, Australia

2. Agukabams (aka Landphoons, Terracanes) From aboriginal roots “agu” (land) and “kabam” (storm) Case studies Case studies Hypothesis Hypothesis Simple model Simple model

3. Case 1: TC Abigail, Feb-March, 2001

7. Best-track winds

8. Case 2: TC John, Jan-Feb 2006

10. Best-track winds

11. Case 3: Tropical Storm Erin, 2007

13. But then, deep in the heart of Oklahoma....

18. Was that a Hurricane? In Oklahoma? Tropical Storm Erin May Have Become A Hurricane... Deep Inland April 12, 2008 at 6:13AM by Chris Mooney

19. Tropical Cyclone Report Tropical Storm Erin (AL052007) 15-17 August 2007 Richard D. Knabb National Hurricane Center 7 April 2008 “Moving northwestward to the south of a large deep-layer ridge over the southern United States, the cyclone became a tropical storm with maximum winds of 35 kt by 1800 UTC 15 August while centered about 180 n mi east of Brownsville..... The center of Erin made landfall at about 1030 UTC that day on San Jose Island, Texas. By that time Erin had weakened to a tropical depression with maximum winds of 30 kt. The depression continued northwestward and inland, and degenerated to a remnant low by 1200 UTC 17 August when it was centered about 50 n mi south of San Angelo, Texas.”

20. “As the surface low moved generally east-northeastward over Oklahoma early on 19 August, the associated thunderstorm activity abruptly increased as the low interacted with an eastward-moving upper-level shortwave trough. During an approximately six-hour period that morning, sustained surface winds as strong as about 50 kt were observed at several locations in western and central Oklahoma, with isolated gusts as strong as about 70 kt. The system’s organization also became dramatically enhanced, with an eye-like feature readily discernible in WSR-88D radar imagery between about 0800 and 1300 UTC as the center of the low passed just north of downtown Oklahoma City. This episode was short-lived, however, and the eye-like feature quickly dissipated after 1300 UTC.”

21. “While the system's structure, particularly its convective organization as seen on radar, resembled and had some characteristics of a tropical storm for a few hours on 19 August, the prevailing view from the Hurricane Specialists at the National Hurricane Center (NHC) is that the system was not a tropical cyclone over Oklahoma.... It is speculated that the upper-level shortwave trough forced the deep convection to increase via upper-level difluence, while briefly superimposed above the surface low that provided a focus for low-level confluence. The upper-level forcing was apparently a dominant mechanism, which is in contrast to tropical cyclones that are maintained primarily by extraction of heat energy from the ocean.... Given all of the considerations described above, the system is simply designated as a “low” by NHC on 19 August.”

22. 12 GMT August 18 2007

23. 18 GMT August 18 2007

24. 00 GMT August 19 2007

25. 06 GMT August 19 2007

26. 12 GMT August 19 2007

28. Hypothesis: Non-baroclinic rejuvenation of tropical cyclones over land can be caused by rapid flux of heat out of hot, sandy soils that have been moistened by the early rains of the system

29. Oklahoma Soils

30. Soil temperature record at Hinton, OK

31. Estimate of Soil Heat Flux

32. 20 cm Soil Temp at Halls Creek during Passage of Abigail

33. A Simple Coupled Soil-Atmosphere Model

34. Coupled Hurricane Intensity Prediction System (CHIPS) Atmospheric Component: –Gradient and hydrostatic balance –Potential radius coordinates give very fine (~ 1 km) resolution in eyewall –Interior structure constrained by assumption of moist adiabatic lapse rates on angular momentum surfaces –Axisymmetric –Entropy defined in PBL and at single level in middle troposphere –Convection based on boundary layer quasi-equilibrium postulate –Surface fluxes by conventional aerodynamic formulae –Thermodynamic inputs: Environmental temperature and soil surface enthalpy

35. Soil Component : –Simple 1-D soil model

36. Boundary condition at top of soil: Lower Boundary Condition: Fixed temperature at 2 m. Thermal conductivity assumed to jump to higher value when soil moisture exceeds a threshold Soil equation solved using 100 layers between the surface and 2 m.

37. Idealized Experiments: Constant translation velocity and environmental temperature, initialization with a warm core vortex with maximum winds of 17 ms -1

38. Sensitivity to Initial Soil Temperature (translation speed = 13 km hr -1 )

39. Sensitivity to thermal diffusivity

40. Sensitivity to translation speed

41. Effect of depletion of water: PBL environmental RH declines from 80% to 0% over 8 days

42. Abigail Hindcast

43. Soil Temperature Evolution

44. John Hindcast

45. Preliminary Hindcast of Erin

46. Summary Tropical cyclones occasionally re-intensify over land with no apparent baroclinic interactions In the cases we examined, the underlying soil was hot and potential intensity based on the temperature of the uppermost 3-50 cm was high

47. Estimates of heat fluxes from the soil, based on soil temperature time series, and results using a simple coupled model support the hypothesis that such events are driven by rapid loss of heat from hot, sandy soils that are wetted by the first rains of the system. We propose to call such systems “agukabams”.