DIAMET. DIAbatic influences on Mesoscale structures in Extratropical storms

Size: px

Start display at page:

Download "DIAMET. DIAbatic influences on Mesoscale structures in Extratropical storms"

Loren Perkins
5 years ago
Views:

1 DIAMET DIAbatic influences on Mesoscale structures in Extratropical storms Geraint Vaughan, Manchester PI John Methven, Reading PI Ian Renfrew, East Anglia PI Doug Parker, Leeds PI

2 DIAMET is part of NERC Storm Risk Mitigation Programme NERC strategic research activity to improve forecasting of storms and their impacts on catchments and coasts 4.9M from DIAMET - Lead PI, Geraint Vaughan, Manchester weather observation, process-level understanding and high- resolution weather prediction TEMPEST - Lead PI, Len Shaffrey, Reading climate DEMON flooding scale) representation of cyclones and stormtracks in climate models and their interaction with climate change - Lead PI, Paul Bates, Bristol improving ability to forecast floods using a cascade of models from NWP rainfall hydrology inundation (street

3 DIAMET Aims To describe the structure, origin and dynamical consequences of heating (condensation/evaporation on the mesoscale e.g., fronts). To make a link between the simulation of mesoscale structure and forecasts of surface weather (high winds and precipitation). To investigate the implications for predictability and the skill of numerical weather prediction Especially convection-resolving forecasts (1-2 days)

Overarching scientific questions for DIAMET What role do diabatic processes play in generating mesoscale potential vorticity (PV) and moisture anomalies in cyclonic storms?

4 Overarching scientific questions for DIAMET What role do diabatic processes play in generating mesoscale potential vorticity (PV) and moisture anomalies in cyclonic storms? What are the consequences of those anomalies for the weather (rainfall and surface winds)? We focus on two key diabatic processes: latent heating/ cooling and air-sea fluxes of heat and moisture.

5 Structure of project WP A. Observation and detailed modelling WP B. Parameterisation of physical processes WP C. Predictability at the mesoscale

6 Structure of project WP A. Obs campaign in 4 phases Sept, Nov/Dec 2011 May, July/Aug 2012 WP B. Convection in cyclones Ocean BL fluxes Phase changes in clouds WP C. Ensembles and data assimilation Balances at small scales Link to precip

7 Structure of project Project lead: Geraint Vaughan Campaign lead: Keith Bower Deliverable leaders: WP A. Doug Parker, Sue Gray, Jeffrey Chagnon, Peter Knippertz, Dave Schultz WP B. Bob Plant Ian Renfrew Tom Choularton WP C. John Methven, Tom Frame, Peter Jan van Leeuwen, Ross Bannister, Stefano Migliorini

8 Structure of project Met Office collaborators WP B. Humphrey Lean OBR: Paul Field WP A. OBR: Roy Kershaw, Phil Brown, Jon Taylor Radar: Malcolm Kitchen WP C. Sue Ballard, Nigel Roberts Ensembles: Richard Swinbank DA: Dale Barker

9 Why consider potential vorticity (PV)? For adiabatic, frictionless flow PV is materially conserved. Balanced flow and stratification associated with PV by inversion. BUT, we are interested in non-conservative effects on scales where balance is not near-geostrophic, so why use PV? Enables us to partition effects on flow by process. Shear, stratification Adiabatic, frictionless Initial condition error dominates large-scale PV and flow Diabatic processes + unresolved turbulence Model error dominates? Heating, turbulent fluxes

10 Tracking non-conservative changes in PV PV tracer diagnostics (originally developed by Bob Plant, Reading) are based on a Lagrangian decomposition of the PV field. Full PV conservation equation can be written: where S i denotes the Lagrangian tendency resulting from one physical process in model Tracers q i are initialised as zero but experience only one of the S i terms as well as being advected by the semi- Lagrangian scheme. each tracer shows accumulated { contribution of one process to PV.

PV TRACER ANALYSIS: T-NAWDEX Pilot case, 24 th Nov 2009 PV tendencies accumulated from 00 UTC on 22 November 2009, UM 12 km run The strip of high PV along the cold front is diabatically generated.

11 PV TRACER ANALYSIS: T-NAWDEX Pilot case, 24 th Nov 2009 PV tendencies accumulated from 00 UTC on 22 November 2009, UM 12 km run The strip of high PV along the cold front is diabatically generated. Main contributions are from boundarylayer heating, cloud microphysics, and convection schemes. Upper-level trough Strip of high PV above surface front The high PV strip is involved in the generation of a largeamplitude gravity wave packet. Source: Jeffrey Chagnon, NCAS-weather

12 Dominant model schemes PV changes PV tracer diagnostics Latent heating and cooling Radiative heating and cooling Mixing Convective parameterization scheme Shortwave radiation Horizontal diffusion Microphysics (grid-scale) Longwave radiation Vertical diffusion Latent heating part of the boundary layer scheme Gravity wave drag parameterization scheme Rebalancing the cloud fields after convection Mixing part of the boundary layer scheme 13

13 Dominant model schemes PV changes PV tracer diagnostics Latent heating and cooling Radiative heating and cooling Mixing Convective parameterization scheme Shortwave radiation Horizontal diffusion Large-scale cloud (grid-scale) Longwave radiation Vertical diffusion Latent heating part of the boundary layer scheme Gravity wave drag parameterization scheme Rebalancing the cloud fields after convection Mixing part of the boundary layer scheme 14

14 Partitioning of diabatic PV among physical processes LW radiation Boundary-lyr heating Large-scale cloud Convection

15 Total diabatic PV in section across tropopause fold ØPositive diabatic PV above (on strat side) of tropopause ØNegative diabatic PV beneath (on trop side) of tropopause ØTropopause elevation not significantly altered by direct diabatic PV modification Chagnon, Gray and Methven (2012)

16 Shading = total diabatic PV anomaly (8.8 km level) Dashed contour = 2PVU for advectiononly PV Solid contour = 2PVU for total PV ØPositive diabatic PV above (on strat side) of tropopause ØNegative diabatic PV beneath (on trop side) of tropopause ØTropopause elevation not significantly altered by direct diabatic PV modification Chagnon, Gray and Methven (2012)

17 Diabatic PV near the tropopause tropopause Chagnon, Gray and Methven (2012)

18 Ramifications for synoptic scale evolution PV conservation equation Using observation of little direct diabatic PV modification at tropopause: However, diabatic effects are felt mainly through induced wind, u i Indirect PV modification by diabatic processes via changes to winds. Links directly with moist baroclinic wave theory (De Vries et al, 2010). Cooperative coupling between diabatic PV in troposphere and Rossby waves along tropopause or lower boundary.

19 Diabatic PV near the tropopause Implications for 1. Stratosphere-troposphere exchange LW cooling max at tropopause diabatic PV dipole but no direct PV change at tropopause + continuous diabatic mass transport downwards across tpp. 2. Theory of moist baroclinic waves growth and propagation

20 Periods of interest DIAMET field campaigns Sep 2011 Frontal waves across UK 25 Nov 15 Dec 2011 Severe cyclone sequence across Scotland (into Jan 2012) PANDOWAE case study focus 11 Sep 2011 Extratropical transition and downstream development 7-10 Oct 2011 Downstream development and extreme precip (N. Alps) 7-8 Nov 2011 Medicane and extreme precip (S. Alps)

21 DIAMET flight summary Date IOP Event 16 Sept 1 Convective band ahead of tropopause trough 20 Sept 2 Weak frontal waves propagating on long trailing cold front 23 Sept 3 Ascent along warm conveyor belt, over-riding a strong frontal wave 26 Nov 4 Surface fluxes 28 Nov 5a Double cold front in Atlantic 29 Nov 5b Vigorous cold front passage over Exeter and Chilbolton 1 Dec 6 Bent back warm front near Shetland + surface fluxes 5 Dec 7 Convection west of Scotland (non-polar low case) 8 Dec 8 Sting jet case over Scotland 12 Dec 9 Warm front case south of Ireland 1-2 May 10 Near-stationary cyclone and flood event (W. England) 9-10 May 11 Developing frontal wave

22 Mesoscale PV structures September IOPs Rossby wave on tropopause ridge enhanced diabatically? w+ - q diab _ Ascent forced ahead of moving tpp trough (IOP1) _ + WCB + - resolved ascent or convective? w+ w+ + q diab Low level frontal wave on trailing cold front (IOP3) -role of diabatically induced flow? + q tpp

23 Eady growth rate index Nov/Dec 2011 Ben Harvey, U. Reading

24 DIAMET IOP-8 Cyclone

DIAMET IOP-8 Cyclone Friedhelm 8 th Dec 2011 Rapid development (deepening 40hPa in 24 hours) High system speed (wave slowing as it matured) Shapiro-Keyser structure at low levels All ingredients for

25 DIAMET IOP-8 Cyclone Friedhelm 8 th Dec 2011 Rapid development (deepening 40hPa in 24 hours) High system speed (wave slowing as it matured) Shapiro-Keyser structure at low levels All ingredients for severe surface winds 165mph (gust) on Cairngorm Considerable damage and disruption Science Questions Were the low-level winds well forecast (structure + strength)? Nature of cloud banding around cyclone centre Role of ice evaporation diabatic cooling enhanced descent and stronger winds at BL top

26 Cyclones that can contain sting jets IR satellite image of a sting jet cyclone (windstorm Gudrun, January 2005). From Laura Baker (2011) PhD Thesis Analysis of several cases indicates that sting jets always occur in Shapiro-Keyser type cyclones

27 Sting jets Conceptual picture Banding in the cloud head From Browning (2004)

28 DIAMET IOP-8 NOAA polar orbiting satellite Visible channel 12:35UT 8/12/11

29 DIAMET IOP-8 NOAA polar orbiting satellite Visible channel 12:35UT 8/12/11

30 Met Office UKV high resolution forecast (1.5km grid-spacing) from 03UT 7/12/11 T+33 forecast was remarkably close to observed structure

31 UKV T+9 forecast Radar image

34 The flight track

35 Equivalent potential temperature (K) Cold front to cyclone centre Dropsondes 1-10 NAE model forecast Cold front Cyclone centre Oscar Martinez-Alvarado

36 Wind speed (m s -1 ) Cold front to cyclone centre Dropsondes 1-10 NAE model forecast Cold front Cyclone centre Oscar Martinez-Alvarado

37 Wind speed (m s -1 ) to SW of cyclone centre (across cold conveyor belt) Dropsondes NAE model forecast Point B Cyclone centre

38 Relative humidity (with respect to ice) to SW of cyclone centre Dropsondes NAE model forecast Point B Cyclone centre

39 Relative humidity (with respect to ice) to SW of cyclone centre cloud bands Dropsondes NAE model forecast Point B Cyclone centre

40 Trajectory analysis 1200 UTC Wind speed (shading) 650-hPa Relative humidity (RH>80% stippled) 850-hPa Equivalent potential temperature Trajectory position (black dots)

41 Trajectory analysis 1200 UTC Pressure (hpa) Height (m) Wind speed (m s -1 ) Relative humidity (%)

42 The sting jet leaving Scotland u 3 Coloured by eastward wind v

43 The sting jet leaving Scotland u 3 Coloured by eastward wind v

44 What did we see at the eye of the cyclone? Q

49 Preliminary results IOP8 DIAMET IOP8 cyclone did contain a sting jet Air descending from tip of cold conveyor belt into hook Observed twice, before and after Scotland? First aircraft observations through a sting jet cyclone Cloud banding around cyclone was not slanting Origin of banding? Q.Is evaporation of ice falling into dry air essential for bringing sting jet winds towards ground?

50 Feature tracking in ensembles Here cyclonic features coloured by max wind speed at 1km within a 300km radius Cyclonic features identified using methodology of Hewson and Titley (2010), Met. Apps.

51 Evolution of one ensemble forecast

52 Example: MOGREPS ensemble for IOP8 cyclone Forecasts initialised 00UT 6/12/11 Full tracks shown from each member Black analysis Green shadow Red - control WPC.1: How much does skill of forecast precipitation in 1.5km model depend on skill of mesoscale feature track (from parent global model)?

53 UKV T+9 forecast Radar image

54 Conclusions Ø Direct diabatic modification of PV at the tropopause is weak Ø Diabatic PV is positive above (above LW cooling at tpp) Ø Diabatic PV is negative below (LW, large-scale rain and convection schemes) Ø Net effect is enhanced PV anomaly pattern at tpp Ø faster growth and slower eastward system speed Ø DIAMET next steps Ø calculate diabatic PV in observed cases to infer contribution of diabatic processes to mesoscale structure Ø Invert diabatic PV tracers to quantify their indirect effect on PV

55 Warm conveyor belt case study Observed 24/11/09 by T-NAWDEX pilot flight DIAMET modelling using UM with PV, humidity and theta tracers (partitioned by model processes) PANDOWAE modelling with COSMO plus sampling physical process tendencies along trajectories Key difference influence of mixing (both parameterised turbulence and numerical)

56 Storm risk mitigation integration Project linking DIAMET, TEMPEST and DEMON (PIs: Len Shaffrey, John Methven and Kevin Hodges) 1. Held metric workshop characteristics of cyclones important to flooding 2. Case study from summer 2007 using climate model down to 1.5km (Met Office Unified Model) 3. Using 5-day accumulated precip (over UK) to sample long HIGEM climate run and pull out related cyclones Characteristics of these cyclones? Analyse them using humidity and PV tracers Are there systematic differences between partition between processes in HIGEM and high-resolution NWP

Diabatic processes and the structure of extratropical cyclones

Geophysical and Nonlinear Fluid Dynamics Seminar AOPP, Oxford, 23 October 2012 Diabatic processes and the structure of extratropical cyclones Oscar Martínez-Alvarado R. Plant, J. Chagnon, S. Gray, J. Methven