1. Tropical Cyclones and Climate Change in a High Resolution General Circulation Model, HiGEM Ray Bell Supervisors: Prof. Pier Luigi Vidale, Dr. Kevin Hodges and Dr. Jane Strachan

2. Introduction Motivation Socio-economic impacts and changing risk with climate change. Impacts on the climate system, removing heat and moisture from the ocean affecting large scale circulation. Research Objectives Investigate the changes in TC activity (location, frequency, intensity, structure and duration) with climate change. Investigate a change of natural variability mechanisms on TC activity e.g. changing ENSO. [Investigate the impact of atmospheric resolution on TC activity with climate change.] [How does TC activity change during a transient forcing vs. stabilised forcing?]

3. What’s expected? How will TC frequency change with climate change? (The number of TCs which form each year) Decrease Stay the same Increase “Warmer SSTs lead to more TCs” “What controls TC frequency?” TC freq = Σ Σ f(SST, dT/dz, mid level humidity, absolute vorticity, vertical wind shear) + Initial disturbances Gray (1968); Emanuel (2005) Space time

4. What’s expected? How will TC intensity change with climate change? (The maximum intensity a TC can reach) Decrease Stay the same Increase TC intensity is governed by its immediate external environment and internal processes TC intensity is hard to simulate in global models (constrained by resolution) - Large regional uncertainties (Knutson et al. 2010) - Can we observe these changes? (Klotzbach and Gray, 2011)

5. Why do we get these changes? TC frequency Decrease globally by 6-34% by 2100 Distribution of SST change. Change in gradients -> Increase in VWS via thermal wind balance (Vecchi and Soden, 2007) Weakening of the tropical circulation. Increase in dry static stability. (Vecchi et al, 2006) TC intensity Decrease of TC frequency mainly from weaker storms. When conditions are favourable for TC development it will be able to utilize the energy available. Supported by theory (Emanuel, 1985) and idealised studies (Shen, 2000) Increase globally by 2-11% by 2100 Knutson et al (2010)

6. Idealised GCM simulations HiGEM UK’s new High-Resolution Global Environmental Model (Shaffrey et al, 2009) 1.25 o x0.83 o, ∆x 50N = 90 km 1/3 o ocean model HiGEM Transient 2% CO 2 /yr 70 yrs HiGEM 1.1 CTRL 150 yrs HiGEM 2xCO 2 30 yrs HiGEM 4xCO 2 30 yrs HiGEM 1.2 CTRL 117 yrs HiGEM CTRL ~9x30 yrs

7. 1) Locate and track all centres of high relative vorticity  35000/yr 2) Apply a 2-day filter to the tracks  8000 storms / yr 3) Analyse vertical structure of storm for evidence of warm-core (tropical storm structure)  120 storms / yr Tracking algorithm (TRACK; Bengstton et al, 2007)

8. Assessing the model At this model resolution we are able to realistically capture location and frequency Strachan et al, (2012) in rev

9. Climate Change Simulations Track density difference 2xCO 2 - CTRL 4xCO 2 - CTRL Stippling if outside 9x30yr CTRL variability Storms/month/10 6 km 2

10. Climate Change Simulations Grey shading is 9x30yr CTRL variability norm pdf Max rel vor Increase in maximum intensity 

11. Large scale forcing NAtl NEPac % change SST VWS pptRH 700 ω 500 TC freq stronger TC freq % change 50 -50 50 -50 Red: 2xCO 2 - CTRL Green: 4xCO 2 - CTRL

12. Conclusion HiGEM realistically captures the geographical location and TC frequency compared to observations. HiGEM simulates a decrease of TC frequency in most regions except for the North Indian Ocean and North Central Pacific region. HiGEM simulates an increase of TC intensity, which only becomes significant in the 4xCO 2 experiment. An increase in VWS in the 4xCO 2 over the North Atlantic spreads to the North East Pacific and decreases TC freq. A weaker Walker circulation suppresses activity in the North West Pacific and enhances activity in the North Central Pacific.

13. Future work Continue Adding HiGEM1.2 onto my current study. Investigate the ENSO relation and different types of El Niño and the impact they have on TC activity. How these change with climate change. Apply my analyses to the different resolution simulation (atmosphere only) Apply my analyses to the transient simulation

15. TC and climate change studies Comparing with, and Different tracking algorithms Model TCs are not exactly comparable to obs TCs Inhomogeneities of obs TCs in different basins and over time. Different models resolutions/ different scenarios - show different parameters to be of importance

16. Late 21st Century projections: “storm-friendly”“storm-hostile” Average of 18 models, Jun-Nov Vecchi and Soden (2007 )

17. Vecchi et al., (2008)

18. Large-scale tropical Atlantic climate changes projected for late 21 st century by CMIP3 models (A1B scenario). Average SST change in MDR is 1.72 o C with warming near 4 o C in the upper troposphere. Knutson et al., (2008) Temp anom

19. Shen (2000) Emanuel Potential intensity

20. Small decrease of TCs. Small increase of major hurricanes TC Hurricanes

21. Increase in maximum intensity 

22. MSLP-10m windspeed relation 5 member ensemble for N512, all 2005 Other models – all years Note: IBTrACS uses 10min winds, models use instantan. 6hourly winds

23. T42 ξ 850 – Reduce noise. Comparison of different spatial resolution data Minimum lifetime of 2 days and no constraint on the minimum displacement distance. Capture more of TC lifecycle Cyclogenesis (0-30 o N over ocean) Coherent vertical structure and warm core Max T63 vor at each level from 850hPa to 250hPa Intensity threshold T63 ξ 850 > 6x10 -5 s -1, ξ 850 – ξ 200 > 6x10 -5 s -1, for at least 1 day (4 x 6hr). Search for warm core between p levels 850-500, 500-200hPa (+ ξ value) Wind speed must attain 20m/s at 850hPa (change in slightly more intense TCs) [att20 dataset] Statistical packages TRACK Hodges (1995); Bengstsson et al. (2007)

24. Understanding natural variability ENSO’s impact on geographical location

25. Understanding natural variability ENSO’s impact on TC frequency

27. Climate Change Simulations Error bars are max and min of 9x30 yr CTRL variability Change in TC frequency

28. Change in SST Zhao et al (2009)

29. AMO ~= AMOC Klotzbach and Gray (2011)

30. Sea Surface Temperature Difference 2xCO 2 - CTRL Sea Surface Temperature Difference (°C) Jul-Oct Tongue of relatively less warm water compared to the rest of the tropics Grave results of TCs in this vicinity (NAtl). Leads to increased vertical wind shear (VWS) via thermal wind balance 4xCO 2 - CTRL

31. Vertical Wind Shear Difference Vertical Wind Shear difference (m/s) Jul-Oct VWS spreads to the NEPac especially in the 4xCO 2 Detrimental affect on TCs. Reduced VWS in CPac favours development Stippling if outside 5x30yr CTRL variability 2xCO 2 - CTRL 4xCO 2 - CTRL

32. Walker Circulation Difference Jul-Oct 0-10N° Weakening of the tropical circulation inline with other studies (Vecchi and Soden, 2007) Favours development in the CPac and reduces TC frequency is the NWPac -ω difference (Pa/s) and divU difference (m/s) 2xCO 2 - CTRL 4xCO 2 - CTRL

33. Change in RH 700 Vecchi et al (2007)

34. Change in –ω 500

35. Change in ppt

36. Large scale tropical change

37. Climate Change Simulations

38. HadGAM – N96. 135km HiGEM - N144. 90km NUGAM - N216. 60km