Geraint Vaughan

Overarching questions for DIAMET What is the role of diabatic processes in generating mesoscale potential vorticity (PV) and moisture anomalies in cyclonic storms? What are the consequences of those anomalies for the weather we experience? We focus on two key diabatic processes: latent heating/cooling and air-sea fluxes of heat and moisture. Poster no.11

Structure of DIAMET WP A. Detailed modelling and measurements WP B. Parameterisations WP C. Predictability Project combines field campaigns, high-resolution modelling, ensemble analysis and data assimilation Poster no.11

Field Measurements. BAe 146 aircraft Chilbolton radar Radiosondes and MST (wind profiling) radar Poster no.20

Field Campaigns in 2011 DateIOPEvent 16 Sept1Convective band ahead of upper- level PV max 20 Sept2Baroclinic waves propagating on long trailing cold front 23 Sept3Ascent in warm conveyor belt 26 Nov4Surface fluxes 28 Nov5aDouble cold front in Atlantic (dropsonde flight) 29 Nov5bCold front passage over Exeter and Chilbolton 1 Dec6Bent back warm front near Shetland + surface fluxes 5 Dec7Organised convection west of Scotland 8 Dec8Bent back warm front: Windstorm over Scotland 12 Dec9Warm front approaching UK from the west

IOP 1, 16 Sept 2011: Convective band 320 K PV 0600 UT 1200 UT

Flight B647 coloured according to altitude (left) and ozone (below) Dots denote dropsondes 320 K PV at Tropopause fold

Cross-section along 52.5 N

MST radar observations These show that the fold was only prominent on the western side of the PV anomaly, and suggest that the rain band was located underneath the fold. fold

Features of this case Convective band, diagnosed as an occlusion on the surface charts, intensified as it passed over Ireland A second band formed later over NE England Band was located beneath an upper-level PV anomaly, but not in a consistent fashion What led to the development of the band? How did diabatic generation of PV in the band affect its development?

IOP 8, 8 Dec 2011: Sting Jet? Surface gusts at 1200, mph Cyclone Friedhelm, with bent- back warm front, passed over Scotland. Strong surface winds caused widespread damage in the Glasgow – Edinburgh region. Aircraft flight measured three cross-sections of the system

1200 IR + 10 m wind (ECMWF) Figure: Eumetrain

Features of this case Strong low-level winds observed in the dry slot of the storm Some evidence for descending air in leg 3 Key questions from this case are: what caused the air to descend? What caused the near-vertical bands of convection wrapping around the low centre, interleaved with stratospheric air? On a larger scale is the question of predictability for this kind of storm, and for the maximum low-level winds within it. Posters no. 17 and 22

Diagnosing mesoscale structures in MetUM simulations of DIAMET IOPs Our aims are to answer the following questions: Which physical processes contribute to the creation of the mesoscale flow anomalies? How do these flow anomalies evolve? / How are they maintained ? What are the consequences of these flow structures? Are they involved in the development of severe weather? Are they modelled accurately? Jeffrey Chagnon, Reading Poster no. 8

EXAMPLE: DIAMET IOP3, 23 September A mesoscale band of heavy precipitation formed along a secondary warm front and moved northwards -The rain band was situated on the northern edge of a diabatic potential vorticity (PV) anomaly -The rain band was maintained by ascent in the flow ahead of the northward moving diabatic PV anomaly Jeffrey Chagnon, Reading LEFT: Partitioning of PV at z = 4.5 km elevation in a 12 km NAE MetUM simulation.

Latent heat of sublimation Specific heat of moist air at constant pressure Mass growth rate by deposition (need to calculate this!) Latent heating rate due to deposition Calculation of latent heating rates from the in-situ microphysics measurements E.g. for deposition growth of ice: Mass growth rates is calculated from the in-situ data in conjunction with an explicit bin microphysics parcel model (ACPIM) developed at the University of Manchester This provides a means of validating (and ultimately improving) calculations of latent heating from simpler bulk microphysics parameterizations such as those used in the WRF model Chris Dearden, Manchester

Diagnosis of diabatic heating rates from microphysical processes in the WRF model (v3.3.1) Region of frontal uplift Ice growth by deposition in the warm sector CondensationDeposition Plots show vertical cross-sections along 51N from simulation of IOP 5 (29 th Nov 2011) Average heating rates between 09:00z and 09:30z for condensation and deposition growth Chris Dearden, Manchester Posters no.5 and 9

Ensemble Prediction of Cyclonic features Met Office cyclone database contains tracked mesoscale features (frontal waves and cyclones) for all MOGREPS15 ensemble forecasts from 2006 to present day. DIAMET WPC1 will use this data to: 1.Examine the statistics of tracked mesoscale features in ensemble forecasts. 2.Quantify predictive skill of features. 3.Quantify the relationship between the skill in forecasting cyclonic features and the skill in forecasting high impact weather. Tom Frame, Reading Poster no.4

First Results Number of features decreases with forecast lead time indicating model dynamics have a smoothing effect. Perturbed ensemble members have more cyclonic features than the control member. Implies uncertainty in forecast associated with dynamical features. N.B. resolution is the same for both. Tom Frame, Reading Poster no.4

Forecast error at convective scales Aim is to investigate sources and characteristics of forecast error at convective scales for use in data assimilation See also posters by Ali Rudd et al. on “Representation of model error in convective-scale ensembles” and by Ross Bannister et al. on “Data assimilation for multi-scale atmospheric flow” Errors in initial and boundary conditions as well as in model formulation accounted for using ensemble prediction techniques 24-member ensemble at 1.5-km horizontal resolution over the southern UK for 20 September 2011 (IOP2) run on the Met Office supercomputer. Initial and boundary conditions were generated using the operational regional version of the Met Office ensemble prediction system (MOGREPS). deterministic truth Stefano Migliorini, Reading Poster no.15

Recent results: representing model error Random variations of parameters in boundary layer and large-scale precipitation schemes in the UM to represent model error (RP scheme) Sensitivity to the parameter ec auto (controlling the conversion of cloud liquid water to rain) on 4 December 2009 at 2300UTC 24-member ensemble spread of different formulations of the RP scheme (Perturbed – Control) ControlPerturbed – Control RMSE rain rate1.5m temperature 10m wind speed No RP scheme (initial condition perturbations only) - Original RP scheme RP scheme with extra parameters - RP scheme with independently varying parameter perturbations Stefano Migliorini, Reading Poster no.15

Conclusions DIAMET is making excellent progress Field campaigns very fortunate with the weather Two more campaigns either in progress or planned Some initial problems on the modelling side now overcome Several good impact activities e.g. national news, videos for schools, EUMETCAL.