PP UTCS Status Report Dmitrii Mironov German Weather

PP UTCS Status Report Dmitrii Mironov German Weather Service, Offenbach am Main, Germany dmitrii. mironov@dwd. de 6 September 2011 COSMO General Meeting, Rome, Italy. 5 -9 September 2011 .

Outline è Overview of project results è Plan for the next COSMO year (September 2011 August 2012) è Brief overview of two projects implemented within the framework of the “Extramurale Forschung” Program of DWD and German Universities COSMO General Meeting, Rome, Italy. 5 -9 September 2011 .

Task ii. a: Matthias Raschendorfer Work Content • Component testing of the COSMO-model turbulence scheme (with particular emphasis on the revised surface-layer transfer scheme) through numerical experiments with COSMO SCM (single-column version of COSMO), using observational and numerical data for verification and forcing

Task ii. c: Matthias Raschendorfer Work Content • Partial reformulation, consolidation and further extension of the current • • • COSMO-model moist turbulence scheme towards a separate realizable scheme for stochastic turbulence (as different from quasi-organised, e. g. convective, motions), including the surface-layer transfer scheme Implementation of the cloud water-cloud ice mixed phase formulation into (the official version of) the COSMO moist turbulence scheme Investigation of the effects of the SGS processes, viz. , stochastic turbulence and quasi-organised convection, on the saturation adjustment (towards a more consistent saturation adjustment procedure that accounts for SGS mixing) Re-organisation of the model code Verification of scale-interaction terms (horizontal-shear and SSO terms already included into an official COSMO-model code) against aircraft data on eddy dissipation rate using the software package mentioned in the previous project report Preparation of detailed documentation

Task ii. a: Results (SCM, surface layer) • An updated COSMO-SC adapted to COSMO version 4. 14 is developed, incl. recent results from Tasks 2. a and 2. c, and is ready for component testing. The SC version as of March 2011 was provided for some COSMO colleagues (e. g. Veniamin Perov). First component tests with COSMO-SC are performed, most notably, in the case of stable stratification of the near surface layer. • A revised formulation of the surface-to-atmosphere transfer scheme is developed (to be implemented later). It is based on a more realistic interpolation of the vertical profile of diffusion coefficient through the transfer layer, incl. the laminar contribution. An improved representation of the diurnal cycle of the near-surface quantities is expected.

Task iic: Results (turbulence scheme) • MODULE “turbdiff” is reformulated in order to use the same module within ICON and COSMO. Modifications include data transfer (e. g. subroutine arguments instead of USE statements), more efficient vectorization, time step handling, options for updated tendency fields or updated fields of prognostic variables. The updated “turbdiff” runs in ICON but is still to be included into the official COSMO code. • An option is introduced to perform the saturation adjustment with due regard for SGS turbulent fluctuations (i. e. to use statistical cloud scheme as a saturation adjustment procedure). Then, the model prognostic variables (T, qv, qc) are in saturation equilibrium with respect to fluctuating fields but may not be in equilibrium with respect to grid-box mean quantities. This option is available in ICON and in the private test version of COSMO but not yet in the official COSMO code. • The development of a “universal” diffusion-equation (tri-diagonal) solver has started (can be used for any passive tracer and for arbitrary vertical grid and boundary conditions, should simplify computations within COSMO). The major development work follows shortly.

Task ii. c: Results (turbulence scheme) • A version of the statistical cloud scheme including the cloud water -cloud ice mixed phase is upgraded (the mixed phase has been tested within earlier COSMO versions by Euripides Avgoustoglou). Principal questions remain considering that cloud ice is generally not in saturation equilibrium. • New version of the turbulence scheme that accounts for the TKE production due to SSO effects has become operational at DWD. Testing further modifications related to the effect of horizontal shear on the TKE production against EDR data of the ACARS archive (using the TMOS package developed within CAT project) is delayed since a follow-up project has not yet started.

Task ii. c: Results (turbulence scheme) • A formulation for the extra TKE source term due to SGS convection based on the Tiedtke cumulus convection scheme is derived. A (very first version of) formulation is implemented into the private test version of COSMO and needs to be verified. It is planned to introduce modifications to the convection scheme in order to improve its coupling with the turbulence scheme (modifications are based on the scale-separation ideas and concern total, i. e. convective and stratiform cloud cover, separate treatment of convective updraugts, environment and downdraughts, and generation of turbulence by coherent convective motions). Further theoretical work has been performed; a first set of governing equations for the new coupled schemes is ready as the basis for further investigation. • The current formulation of the “circulation term” in the TKE equation is being modified to account for coherent thermal circulations generated by SGS slopes. The work has just started. • For further details, see the Task ii. c report at the COWMO web page and Matthias’ presentation at the WG 3 a/UTCS session

DWD Matthias Raschendorfer COSMO Rome 2011

including horizontal shear – and SSOproduction referen ce pot. temperature [K] including horizontal shear –, SSO- and convective production DWD Matthias Raschendorfer COSMO Rome 2011

Task ii. b: Veniamin Perov Work Content • Generalization of non-local formulation of the turbulence length scale to account for moist processes (effect on phase changes on the scale of turbulence) • Comprehensive testing of a generalized formulation through single-column and 3 D experiments • Verification of results in different weather situations • Preparation of documentation Results • An algorithm for computing the non-local turbulence length scale is generalized to account for the effect of cloud condensate • The new algorithm is included into the module TURBDIFF of COSMO-RU • Numerical experiments show a satisfactory performance of the new length-scale formulation (overall, turbulence as well as mixing length and buoyancy fluxes are more intense in the case of “moist” mixing length) • Numerical experiments are performed with 3 -D COSMO model for summer days, results are compared with the reference version of COSMO and with the radiosonde and profiler data

Task ii. b: Results Vertical profiles of potential temperature computed with the current (blue) and new (red) formulations for the turbulence length scale. Observational data are shown with the black curve.

Task ii. b: Results (cont’d) T 2 m difference (non-local L local L), left panel, and total cloud cover, right panel, for the costal region of Black Sea 17 July 2009

Task iii: Euripides Avgoustoglou Work Content • Systematic evaluation of the performance of the SGS cloud schemes, most notably the version with the water-ice mixed phase, against satellite data (also with data on T 2 m, Td 2 m, and radiation budget at the surface) • Critical assessment of the suitability of the statistical cloud scheme for use in radiation calculations Results (see Euripides’ presentation, PPT file available via COSMO web page) • Different version of cloud schemes are tested against satellite and in-situ observational data, including the default relative-humidity scheme and the statistical cloud scheme with the cloud water-cloud ice mixed phase • Total cloud cover compares decently with the satellite data for both schemes • Medium and low clouds: the PH scheme outperforms the RH scheme in areas with low cloudiness and the RH scheme fares better in more cloudy areas • Minimum T 2 m: the mixed-phase scheme performs better; maximum T 2 m: in favour or RH scheme • The mixed-phase scheme is recommended as an option within COSMO, although a systematic investigation will be of value to fully evaluate the mixedphase scheme

Task iii: Results (cont’d) Deg C 28 -April-2011 T 2 m_min Station # Minimum T 2 m 28 April 2011. The vast majority of values from mixed-phase scheme runs (R_mix_0. 5_4. 0 and R_mix_0. 5_2. 0) are closer to observations than the corresponding values of the default relative-humidity scheme of COSMO model (R_rh)

Task iii: Results (cont’d) 28 -April-2011 T 2 m_max Deg C Station # Maximum T 2 m 28 April 2011. The values from the default RH scheme of COSMO model runs (R_rh) are closer to observations than the corresponding values from the MP scheme runs (R_mix_0. 5_4. 0 and R_mix_0. 5_2. 0) but the situation is more balanced than for minimum T 2 m.

28 April 2011 02: 45 UTC Low Cloud Cover (LCC) Meteosat/Synesat % R_rh R_mix_0. 5_4. 0 Comparison of Low Cloud Cover (LCC) against the corresponding MSG satellite image (Meteosat/Synesat) manipulated with Synesat software available at HNMS. The overall trend (especially in the encircled areas) is that the mixed -phase scheme (R_mix_0. 5_4. 0) fares better than the relative-humidity scheme (R_rh) in reference to Meteosat satellite image.

Task i: Ekaterina Machulskaya and Dmitrii Mironov Work Content • Consolidation of a second-order closure scheme of a moist PBL, including transport equations for the TKE and for the scalar variances (liquid water potential temperature, total water specific humidity) and a sub-grid scale statistical cloud scheme for non-precipitating clouds • Implementation of the new closure scheme into the COSMO model • Testing through numerical experiments within the framework of the fullfledged three-dimensional COSMO model, including the assimilation cycle, and through single-column experiments as needed • Analysis of results from numerical experiments, comparison of results with observational and numerical data, tuning of the new scheme as needed • Preparation of documentation

Task i: Results • The TKE-Scalar Variance scheme is implemented into the COSMO model (new features: prognostic equations for the scalar variances with third-order transport terms and non-gradient corrections to the fluxes of scalar quantities) • The new scheme is being tested through parallel experiments with COSMO-EU (ca. 7 km mesh size) and COSMO-DE (ca. 2. 8 km) model configurations using COSMO version 4. 18 operational at DWD since May 2011. Results from both COSMO-EU and COSMO -DE show sensitivity to the modifications introduced (e. g. the surface-layer in summer seems to be moister). Verification results and conclusive statements as to the effect of the new scheme on the overall COSMO performance will be reported later. • Preparation of documentation in progress • Parallel experiments are running slower than expected, coding errors (our fault!), work somewhat delayed

Task i: Results (cont’d) Temperature gradient difference (Exp – Routine) in the lower part of PBL 05. 07. 2011 (upper panel) and 10. 07. 2011 (low panel)

Task i: Results (cont’d) Coupling of the TKE-Scalar Variance Scheme with Tiled Surface Scheme (most important for stably stratified PBL) Motivation • Results of Patrick Volker (reported at COSMO GM 2010) and Mironov and Sullivan (2010) • Among other things, (i) minimum eddy diffusion coefficients should be (drastically) reduced but (ii) turbulence in shallow and strongly stable PBL should be maintained A way to go • Tile approach to account for the effect of surface heterogeneity on mixing in the SBL

LES of SBL over Heterogeneous Surfaces (Mironov and Sullivan 2010) Blue – homogeneous SBL, red – heterogeneous SBL. • Enhanced mixing is due to increased temperature variance close to the surface (see Mironov and Sullivan 2010, for details) • For want of a more elegant theory, use tile approach where the number of tiles should not be large (otherwise computationally expensive) but account for tiles with a maximum difference in terms of thermal inertia • SGS water is crucial

TKE-Scalar Variance Scheme and Tiled Surface Scheme • Tile approach where different tiles have different surface temperature (the mosaic scheme coded by Felix Ament was used as starting point) • Surface fluxes are computed as weighted means of fluxes over individual tiles • Transport (prognostic) equations for TKE and for the scalar variances including third-order transport of scalar variance • < ’ 2>, <w’ ’ 2>, <q’ 2> and <w’q’ 2> determined with tiled scheme (non-zero at the surface!) are used as lower boundary conditions for scalar variances First step: testing tiled surface scheme with due regard for SGS water

Sensitivity Experiment • COSMO-EU, the lake parameterisation scheme FLake is used to model inland water bodies • External-parameter fields of lake fraction and lake depth are generated using the updated lake-depth data set (Kourzeneva 2010) and the software package by Hermann Asensio • For grid boxes with 0. 05<FR_LAKE<0. 5 (SGS water), the surface fluxes are computed as weighted mean over two tiles, water and land • All SGS inland water bodies are 10 m deep (externalparameter software should be modified to use the actual depth where available)

Task i: Results (tiled scheme) The lake-fraction external-parameter field based on the lake-depth data from Kourzeneva (2010) and Glob. Cover physiographic data. The horizontal size of the COSMO-model grid is ca. 7 km.

Task i: Results (tiled scheme, cont’d) Difference in surface temperature between experiment with tile approach (mean over two tiles) and control experiment (no tiles) night, 30. 04. 2011 00 UTC Warming due to SGS lakes day, 30. 04. 2011 12 UTC Cooling due to SGS lakes

Task i: Results (tiled scheme, cont’d) Control run (no tiles) T 2 m warm bias is reduced Experiment (two tiles)

Plans for the Next COSMO Year Task i. Ekaterina Machulskaya and Dmitrii Mironov • Comprehensive testing of the new TKE-Scalar Variance scheme (including transport equations for the TKE and for the scalar variances, a sub-grid scale statistical cloud scheme for nonprecipitating clouds, and non-local formulation for the turbulence length/time scale) through numerical experiments within the fullfledged three-dimensional COSMO model • Analysis of results from numerical experiments, verification of numerical results against observational data • Full coupling of the new scheme with the tiled surface scheme (tiled scheme c/o Ekaterina Machulskaya and Jürgen Helmert), tuning of the coupled schemes • Preparation of documentation

Plan for the Next COSMO Year (cont’d) Task ii. a. Matthias Raschendorfer • Detailed investigation of the problems with diurnal cycle of the near surface quantities using measurements and COSMO-SC • Implementation and verification of a revised surface-layer transfer formulation In the more distant future • Implementation of a canopy/skin layer Task ii. c. Matthias Raschendorfer • • • Generation of an official code valid for COSMO, SCM and ICON Reformulation of the positive definite, semi implicit solver of the TKE equation Implementation of a universal subroutine to solve the implicit vertical diffusion equation Verification of the alternative expression of moist corrections Verification of the implemented scale interaction terms (in particular, against ACARS EDR data using the TMOS package for Turbulence Model Output Statistics) • Reformulation of the current “circulation term” to be a thermal SSO source term for TKE In the more distant future • Implementation of SSO/roughness layer terms based on an extended boundary-layer approximation • Implementation of a scale separated mass-flux convection interacting with turbulence and providing volume fractions of convective sub-domains

Plan for the Next COSMO Year (cont’d) Task ii. b. Veniamin Perov • The work plan for the next COSMO year is being discussed, it will be sent to the COSMO governing bodies shortly after GM • Proposal for the more distant future (beyond the PP UTCS time span) is drafted by Veniamin Task iii. Euripides Avgoustoglou • Evaluation of the performance of the SGS cloud schemes, the statistical cloud scheme with the water-ice mixed phase in particular, against satellite data and data from measurements in the surface air layer • Critical assessment of the suitability of the statistical cloud scheme for use in radiation calculations, recommendations towards the use of various SGS cloud schemes within COSMO model

“Extramurale Forschung” Program of DWD and German Universities Rieke Heinze, Siegfried Raasch, University of Hannover • “High-resolution large-eddy simulation of atmospheric boundary -layer turbulence - contribution to the improvement of turbulence parameterisations through systematic study of higher-order statistical moments and their budgets” Richard Keane, George Craig, Christian Keil, University of Munich • “Development and testing of a scale independent convection parameterisation for ICON”

High-Resolution Large-Eddy Simulation of Atmospheric Boundary-Layer Turbulence Motivation • In spite of numerous LES of PBL turbulence, no comprehensive analysis of the second-moment budgets for cloudy boundary layers • Budgets are required to (i) estimate relative importance of the various terms, (ii) test and further develop turbulence parameterizations, (iii) determine disposable parameters Results • Budgets for shallow cumulus case are estimated on the basis of very high resolution LES • First results for stratocumulus case are obtained Outlook • Sc-Cu transition case • Analysis of pressure-scalar and pressure-velocity covariances possible with observational data!) (not

BOMEX − TKE and its budget

BOMEX − variance of total water content and its budget

BOMEX − temperature flux and its budget !!!

Development and testing of a scale independent convection parameterisation for ICON Motivation • Improve representation of convection in NWP models (meso-scale organisation, momentum transport) • Account for a highly variable (stochastic) convective activity within grid cells Results • The Plant-Craig stochastic convection scheme is implemented into COSMO • The COSMO version of the scheme is adapted to become “GCMindependent” • The GCM-independent version of the scheme is implemented into ICON, tests are performed (idealised cases, aquaplanet) Outlook • Improve the scheme computational efficiency (ongoing) • Perform comprehensive tests within ICON

Rank Probability Scores for rainfall for a (UKMO) MOGREPS forecast PCCS_CTL: (standard) Gregory-Rowntree scheme PCCS_EXP: Plant-Craig scheme The RPS compares the distribution of the ensemble with that of the observations – a lower score is better!

Accumulated convective rainfall in the aquaplanet experiment in ICON: comparison between the Plant-Craig scheme (top) and the (standard) Tiedtke-Bechtold scheme (bottom)

Example rainfall snapshots for three different convection schemes in COSMO The rainfall variability is closer to that of reality with the Plant-Craig scheme, as compared to two different conventional schemes.

Summary è The project will be completed in September 2012 è The scope of work beyond 2012 depends on results (testing within COSMO may require more time than expected, operational implementation may occur after the project is formally completed, fine tuning may be needed) è Future work should be organised within WG 3, new priority projects may be proposed (Matthias, Veniamin) è COSMO folks are strongly encouraged to use results from EMF Projects! COSMO General Meeting, Rome, Italy. 5 -9 September 2011 .

Thank you for your kind attention! Acknowledgements: Peter Bechtold, Vittorio Canuto, Sergey Danilov, Evgeni Fedorovich, Jochen Förstner, Jean-Francois Geleyn, Vladimir Gryanik, Thomas Hanisch, Donald Lenschow, Chin. Hoh Moeng, Ned Patton, Pier Siebesma, Peter Sulluvan, Jeffrey Weil, Jun-Ichi Yano COSMO General Meeting, Rome, Italy. 5 -9 September 2011 .

Comparison of rainfall variability from Plant-Craig with that predicted by a CRM-based theory, for an idealised numerical experiment. This is shown at four different scales, marked by the number in km.

Task ii. b (V. Perov): Results (2011) 700 Vertical profiles of temperature in the boundary layer, for Vertical of Potential Temperatures Moscow, 12 h UTC, 12 h 17. 07. 2009 Dolgoprudnyj, UTC, 17. 09. 600 2500 2000 COSMO New Height, м 400 1500 Measurements from Profiler MPT -5 COSMO- REF 300 200 1000 COSMO reference Зонд COSMO NL 500 0 298 299 300 301 302 303 304 Potential Temperature, К 0 Vertical profiles temperature computed with 17 18 of (potential) 19 20 21 22 the current 23 (blue) and 24 new (red 25 – left panel, green – right panel) formulations for the turbulence length scale. Observational data Temperature, grad С are shown with the black (left) and red (right) curves. 26

Task ii. b (V. Perov): Results (cont’d, 2011) T 2 m difference (New L – Ref) for Moscow area, 17. 07. 2009 12 UTC

Task ii. b (V. Perov): Results (2012) T 2 m difference (New L – Ref) for Moscow area 17 July 2009 12 UTC

Old Stuff

Tiled TKE-Temperature Variance Model: Results Blue – homogeneous SBL, red – heterogeneous SBL.

Task iv: Work did not start Work Content • Comparison of the cloud condensate diagnosed by the sub-grid scale cloud schemes (statistical and relative-humidity) and by the grid-scale saturation adjustment procedure (following the prognostics equations for the cloud water and cloud ice) Expected Outcome • A prerequisite for improving/tuning the statistical SGS cloud scheme and for eventually improving the treatment of the could water and of the cloud ice with due regard for the sub-grid scale processes Martin Köhler (DWD, FE 14) plans to perform this work within the framework of ICON, COSMO will benefit

Problems Encountered • Problems of scientific character: nothing unexpected in view of the innovative nature of the project, i. e. some risk is inevitable and must be assumed • Communication: considering high complexity of the problem, a remote mode operation is often inefficient (e. g. the project leader cannot keep an eye on the details of the implementation of all tasks)