PP COLOBOC Final Report Jean-Marie Bettems / MeteoSwiss Juergen Helmert / DWD COSMO GM Roma, September 7 th, 2011

2 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COnsolidation of LOwer BOundary Condition COSMO priority project, – Goals Facilitate access to soil/surface observations Consolidate tools of general interest: externalized TERRA module software for generation of external parameters Consolidate and extend external parameters database Find and validate an optimal configuration of TERRA Revision of snow analysis and snow model Deployment of urban module developed in Switzerland Consolidate parameterization of land surface heterogeneity

3 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Three years later… What has been achieved? Do we need a COCOLOBOC ?

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COLOBOC, task 0 Observation sets for model validation and development New instruments at Payerne Measurement of 10m, in activity since spring Soil moisture and temperature kept after end of SwissSMEX Data exchange action within SRNWP

5 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Goal: convenient access to recent operational high quality observations, representative for different climate and different type of soils Action started in October 2009 with an official letter sent to a selection of NMS directors. Agreement within the C-SRNWP Programme of EUMETNET : “I would like to emphasise that the data would be provided ONLY for the C-SRNWP members (particularly to the Expert Team members) with a significant time-lag after the observational time and for research purposes only.” Resources from DWD (C.Heret, F.Beyrich), HNMS (T.Andreadis) and MCH (JM.Bettems) as contribution to the COLOBOC project. Support from C-SRNWP coordinator (András Horányi) and from chair of SRNWP ET Soil and Surface (JF.Mahfouf) Data exchange action Background

6 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Data exchange action Data sets Access from COSMO web, password protected Currently 7 sites, data from , in a common ASCII format Soil, surface and BL observations Agreement for two new sites: Debrecen (Hu), Valdai (Ru) Valdai Debrecen

7 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma #SRNWP data exchange action #Cabauw °N 4.927°E : :00 UTC DATE;P0;Q_P0;RAIN;Q_RAIN;TAIR002;Q_TAIR002;TD002;Q_TD002;TAIR010;Q_TAIR010;TD010;Q_TD010;WSPEED010;Q_WSPEE D010;WDIR010;Q_WDIR010;TAIR020;Q_TAIR020;TD020;Q_TD020;WSPEED020;Q_WSPEED020;WDIR020;Q_WDIR020;TAIR040;Q_ TAIR040;TD040;Q_TD040;WSPEED040;Q_WSPEED040;WDIR040;Q_WDIR040;TAIR080;Q_TAIR080;TD080;Q_TD080;WSPEED080;Q _WSPEED080;WDIR080;Q_WDIR080;TAIR140;Q_TAIR140;TD140;Q_TD140;WSPEED140;Q_WSPEED140;WDIR140;Q_WDIR140;TAI R200;Q_TAIR200;TD200;Q_TD200;WSPEED200;Q_WSPEED200;WDIR200;Q_WDIR200;RSWD;Q_RSWD;RSWU;Q_RSWU;RLWD;Q_ RLWD;RLWU;Q_RLWU;CLC;Q_CLC;USTAR;Q_USTAR;MOM;Q_MOM;HS;Q_HS;LE;Q_LE;TSOIL000;Q_TSOIL000;TSOIL002;Q_TS OIL002;TSOIL004;Q_TSOIL004;TSOIL006;Q_TSOIL006;TSOIL008;Q_TSOIL008;TSOIL012;Q_TSOIL012;TSOIL020;Q_TSOIL020;TSO IL030;Q_TSOIL030;TSOIL050;Q_TSOIL050;MSOIL008;Q_MSOIL008;G00;Q_G00;G005;Q_G005;G010;Q_G010 yyyyMMddhhmm;hPa;-;mm;-;°C;-;°C;-;°C;-;°C;-;ms-1;-;degree;-;°C;-;°C;-;ms-1;-;degree;-;°C;-;°C;-;ms-1;-;degree;-;°C;-;°C;-;ms-1;- ;degree;-;°C;-;°C;-;ms-1;-;degree;-;°C;-;°C;-;ms-1;-;degree;-;Wm-2;-;Wm-2;-;Wm-2;-;Wm-2;-;%;-;ms-1;-;Nm-2;-;Wm-2;-;Wm-2;-;°C;-;°C;- ;°C;-;°C;-;°C;-;°C;-;°C;-;°C;-;°C;-;-;-;Wm-2;-;Wm-2;-;Wm-2; ; ;g;0.00;g;11.00;g;4.70;g;11.60;g;4.70;g;4.62;g;61.50;g;11.70;g;4.50;g;5.17;g;63.40;g;11.80;g;4.50;g;7.17;g;66.04;g;11. 90;g;4.00;g;9.65;g;70.25;g;11.90;g;3.90;g;12.17;g;73.79;g;12.50;g;2.60;g;14.21;g;82.79;g;0.00;g;0.00;g;295.30;g;359.90;g; ;m;0.32;g;0.13;g;-65.04;g;17.79;g;12.59;g;12.70;g;13.41;g; ;m;13.53;g;13.78;g;13.74;g;13.51;g;12.37;g;0.26;g;-8.43;g;- 8.38;g;-45.50;g ; ;g;0.00;g;11.00;g;4.90;g;11.60;g;4.80;g;4.02;g;61.30;g;11.70;g;4.70;g;4.99;g;63.50;g;11.70;g;4.60;g;6.84;g;66.75;g;11. 80;g;4.10;g;9.45;g;70.34;g;11.80;g;4.00;g;11.97;g;74.18;g;12.50;g;2.60;g;14.04;g;84.06;g;0.00;g;0.00;g;290.45;g;356.64;g; ;m;0.25;g;0.08;g;-46.81;g;12.29;g;12.55;g;12.66;g;13.35;g; ;m;13.50;g;13.74;g;13.72;g;13.51;g;12.37;g;0.26;g;-8.43;g;- 8.36;g;-45.58;g Monthly CSV tables, for each site, with all parameters, incl. qc when available Missing values are flagged with CAB_ txt.gz Data exchange action Data sets – Examples

8 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma LIN Availability [%] Data exchange action Parameters availability, Lindenberg, status

9 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma SPC Availability [%] Data exchange action Parameters availability, Toulouse, status

10 PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Data exchange action Outlook Promote usage → extend default set of COSMO model meteograms (WG5) → integrate in soil & BL developments (WG3a/b) → routine inter-comparison of soil and surface fluxes (WG5) → promote through SRNWP expert team soil&surface (JF.Mahfouf) Make this effort sustainable → transfer responsibility for this actions to the new COSMO WG3b → F.Beyrich / DWD has agreed ”… continuation of the management of the data streams for the data pool as well as long as this is possible with reasonable efforts.” Open this initiative to academic institutions ? → this would drastically extend the set of potential users → … but a re-formulation of the agreement is necessary

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COLOBOC, task 3 Consolidate external parameters data set. Task 3.1: Document available data Task 3.2: Consolidate and extend database Add seasonal climatology of vegetation characteristics. Add parameters for subscale orography. Add parameters for orographic radiation correction. Add parameters for urban fraction. Add parameters for deep soil temperature. Add parameters for FLake. Provide alternative sets of soil types in Europe. Address the feasibility of deriving surface characteristics with known deficiencies (e.g. solar albedo) using alternative data sources or methods.

Overall description of the External Parameter System external parameters on target grid Orography Soil data Land use

ksh scripts Fortran programs global icosahedral grid GME ksh scripts Fortran programs limited area domain COSMO current status external parameter generation new library “EXTPAR” external parameter generation ksh scripts Fortran programs (Fortran 2003/ Fortran 95) user defined grid Icon GME COSMO

Consolidation and extension of database (COLBOC Task 3) : Additional external parameters  external parameter fields for the COSMO model, total 15 fields  planned extensions for the COSMO model, total 30 fields HH, FR_LAND, SOILTYP, Z0, SSO_STDH, SSO_GAMMA, SSO_THETA, SSO_SIGMA, PLCOV_MN, PLCOV_MX, LAI_MN, LAI_MX, FOREST_D, FOREST_E, ROOTDP HH, FR_LAND, SOILTYP, Z0, SSO_STDH, SSO_GAMMA, SSO_THETA, SSO_SIGMA, PLCOV_MX, LAI_MX, FOREST_D, FOREST_E, ROOTDP NDVI, NDVI_MX, NDVI_RATIO, AER_BC, AER_DUST, AER_SO4, AER_SS, T2M_CL, URBAN, RSMIN, EMISS_RAD, FR_LAKE, DEPTH_LK, SLOPE_ASP, SLOPE_ANG, HORIZON, SKYVIEW

Currently used raw data for external parameters GLOBE DSMW GLC2000, GlobCover2009

 NDVI (SEAWiFS)  climatology of near surface temperature (CRU)  climatology of aerosol optical thickness (I. Tegen)  lake depth database (30’’ gridded global field, Eakterina Kourzeneva: DWD, RSHU, MeteoFrance) Additional raw datasets

Uncertainties The external parameters are only as good as the available raw data sets. GLC2000 land use classes (currently used) Globcover 2009 GLCC USGS land use / land cover system

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COLOBOC, task 4 Revision of TERRA and the associated look-up tables. Gather and integrate tested developments of TERRA in official COSMO code - V /11/30 Define a consolidated set of look-up tables – EXTPAR (ECOCLIMAP, GLOBCOVER) Find and validate an ‘optimal’ reference configuration of TERRA, using the external parameters developed within task 3.2 with the consolidated look-up tables Benefit: > Unification of physics in TERRA standalone, GME, and COSMO > Appropriate configuration for all climate regions

Long-term trend T2M-Bias

Aerosol climatology Emissivity Vegetation climatology (LAI, PLCOV) Minimum stomatal resistance (GLC2000, ECOCLIMAP) Vegetation type albedo Non-uniform root distribution GME configuration

External Parameters GME

Albedo (snow free): Vegetation albedo ROUTI

EXP 7750 Masson (2003) - Ecoclimap Albedo (snow free): Vegetation albedo EXP

Europe East Sibiria North America EXP 7750 ECOCLIMAP+ALBEDO UTC i192f Tropics Africa Tropics America Global Verification SK: -6.03SK: -5.68SK: 1.02 SK: -11.7SK: -6.36SK: SK: -8.00SK: -9.8SK: SK: 1.99 SK: 2.62 SK: 0.37 SK: 3.64 SK: 5.78SK: 0.78

PRS/LAI ratio and vegetation type albedo have strong impact on the split-up of fluxes Positive impact on T_2M and TD_2M in temperate zones Neutral or small negative impact in tropics Since March 2011: Soil moisture analysis (SMA) in GME Conclusions of GME EXP 7750

Bias and Stdev T2m over Europe, avg(12, 15 GMT) Soil moisture content over Europe, top layers Soil moisture content over Europe, bottom layers M. Lange Europe T2m Bias decreased by SMA Soil moisture increase during summer! Running time (2009)

GME-SMA EXP 109/ Experiments SMA-Monitoring (M. Lange)

Validation Europe T2M Bias and StdvSMA Incr. Mean and Stdv bottom layers SMA Incr. Mean and Stdv top layers

Validation Europe T2M Bias and StdvSMA Incr. Mean and Stdv bottom layers Soil moisture W_SO (mm) top layers

soil moisture: level 4 Validation

Verification EUR T2M PS TD2MD TMINTMAX DD FF DD TMIN TMAX VW BIAS RMSE

COSMO Experiments

History CLM Study (ETH)

History CLM Study (ETH)

Validation AOD July 2010 Tegen, 1997 Tanre,1984

Aerosol climatology – could explain increase in T2M-Bias Emissivity Vegetation climatology (LAI, PLCOV) GLOBCOVER 2009 land use data Minimum stomatal resistance (ECOCLIMAP) Vegetation type albedo Summer period : Non-uniform root distribution Ground water with upward diffusion Soil moisture dependent heat conductivity Start with new model initialisation (8310 ROUTI config CTRL) Tested adaptions in int2lm and TERRA EXP 8309/8310

Verification EXP vs CTRL T2M PS TD2MD TMINTMAX DD FF DD TMIN TMAX VW BIAS RMSE 8309/ UTC

Verification EXP vs ROUTI T2M PS TD2MD TMINTMAX DD FF DD TMIN TMAX VW BIAS RMSE 8309/ROUTI 00 UTC

Results from GME experiments suggest that soil moisture analysis (SMA) has a large impact on the results Difficulties for parameterization changes to outperform SMA Impact of model initialisation is important Improved post-processing required to avoid experiment initialisation from large-scale models Benefit: Unified land-surface scheme for global, regional, and stand-alone application Conclusions and Outlook

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COLOBOC, task 5.1 Verify and consolidate the new multi-layer snow model Heat conduction Melting when snow surface temperature > 0°C or when soil surface temperature > 0°C Heat conduction Phase changes, liquid water transport Gravitational compaction + metamorphosis Solar radiation penetration 1 layer Arbitrary number of layers Heat conduction: implicit Latent heat and solar radiation: source terms Numerical schemes Implemented processes Single layerMulti layers

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COLOBOC, task 5.1 Station at 2450m on the southern slope of the Alps Model at 2350m Dh = -100m

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Review – COLOBOC, task 5.2 Improved snow analysis MeteoSwiss has extended the DWD snow analysis by introducing a MSG derived snow mask and by tuning the Cressman analysis of snow depth observations. Unified sfcana software is available. Composite fractional snow cover

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COLOBOC, task 7 Parameterization of land surface heterogeneity by the tile/mosaic approach.

Tiles – Motivation Observed 5cm deep soil temperature Start of forecast Observed 2m temperature Start of forecast By any partial snow cover the soil temperature can not exceed 0ºC (equivalent to instantaneous horizontal heat transfer) One profile of temperature and moisture in the soil for snow covered and snow-free parts T_SO_3, 3 days forecastT2M, 3 days forecast

Tiles – Implementation Either mosaic or tile approach in the version 4.16, with a user-specified number of tiles or sub-grid mosaic pixels. surface type 1 surface type 2 surface type 3 surface type 4 snow- free snow ns_snow ns_stat User-defined parameters: ns_stat: number of static surface types ns_snow: number of those static surface types which can have snow as a separate tile snow fraction for each of these types is recomputed at each time step, i.e. dynamic tiles

Tiles – Example of application Grid-scale lakes (lake fraction > 0.5) are treated within the COSMO with FLake model; there are, however, numerous subgrid-scale lakes (lake fraction < 0.5) which are ignored lake fraction

Tiles – Example of application Difference in surface temperature between COSMO-EU experiment with tile approach (lake/no lake) and control experiment UTC

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Summary of successes and misses

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Priority project COLOBOC Final status (ok, finalized till end 2011, not all goals reached)  T0. Observation sets for model validation and development > new instruments at Payerne > SRNWP data pool action, with 9 sites > coordination transferred to WG3b, resources guaranteed  T1. Externalized TERRA module > consolidated package and documentation available on COSMO web site (based on COSMO 4.13) > not supported, open question in WG3b  T2. Software for generation of external parameters > consolidated software EXTPAR v1.5 > software and documentation are available (software available at DWD, soon on COSMO web site) > web portal soon operational  T3. External parameters > Documentation of data sets available > Consolidated data set, with 15 new parameters  T4. TERRA improvements > Implementation of TERRA developments into COSMO V /11/30 > Definition of the consolidated set of look-up tables in EXTPAR > Experiments at MeteoSwiss and DWD showed mixed, neutral impact on verification scores > Physically more realistic, basis for further developments

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma Priority project COLOBOC Final status (ok, finalized till end 2011, not all goals reached)  T5.1. Improved snow model > revised parameterization of snow albedo (incl. ageing) > revised parameterization of partial snow cover (height instead of snow water equivalent based) > proposition for a new parameterization of old snow density > new multi-layers snow model implemented in COSMO 4.11 > still some stability issues for very shallow snow layers, corrections being investigated > official release, tested and documented, planned for Q  T5.2. Snow analysis > unified snow analysis software (MCH modifications merged in official sfcana) > altitudinal interpolation not available, temporal stability of analysis may be an issue  T6. Urban module BEP > code and documentation available at MeteoSwiss (based on COSMO 4.7) > not supported, open question in WG3b  T7. Parameterization of surface heterogeneities with the tile/mosaic approach > tile and mosaic implemented in COSMO 4.16 > user configurable, dynamic tiles for snow supporting surface types > consolidated and official release planned for Q2 2012

PP COLOBOC, Final Report | COSMO GM, 5-8 September 2011, Roma COSMO GM Offenbach | postprocessing tool THE END