:

General Observation Period (GOP) 2007General Observation Period (GOP) 2007General Observation Period (GOP) 2007Concepts and first resultsConcepts and first resultsConcepts and first results

M i M h1 S C ll1 Th t R i h dt1 Ch i t h S lb h1 d GOP TMario Mech1, Susanne Crewell1, Thorsten Reinhardt1, Christoph Selbach1, and GOP Teampmech@meteo uni-koeln de crewell@meteo uni-koeln de treinhar@meteo uni-koeln de cselbach@meteo uni-koeln de http://gop meteo uni-mech@meteo.uni-koeln.de, crewell@meteo.uni-koeln.de, treinhar@meteo.uni-koeln.de, cselbach@meteo.uni-koeln.de http://gop.meteo.uni-

k l dkoeln.de1 IGMK: University of Cologne Institute for Geophysics and Meteorology Kerpener Str 13 50923 Cologne Germany1 IGMK: University of Cologne, Institute for Geophysics and Meteorology, Kerpener Str. 13, 50923 Cologne, Germany

GOP OutlineGOP Outline

The main goal of the General Observation Period (GOP) within the German instrumentation by gathering routine measurements normally not available to theThe main goal of the General Observation Period (GOP) within the GermanR h F d ti ’ P i it P Q tit ti P i it ti F ti

instrumentation by gathering routine measurements normally not available to thei tifi it F h b t ti / di t d b tiResearch Foundation’s Priority Program on Quantitative Precipitation Forecasting scientific community. Focus has been put on continuous/coordinated observations

(SPP-PQP) is to gather a comprehensive data set suitable for testing hypotheses using existing instrumentations which are suitable for statistical evaluation and on( ) g p g ypand new modeling techniques developed within the SPP-PQP The GOP

g gmeasurements which are available in near real-time to enable a timely use withinand new modeling techniques developed within the SPP PQP. The GOP

encompasses the Convectively and Orographically induced Precipitation Studymeasurements, which are available in near real time to enable a timely use withinthe PQP The GOP group performed a rigorous quality control cross checking andencompasses the Convectively and Orographically induced Precipitation Study the PQP. The GOP group performed a rigorous quality control, cross-checking, and

COPS performed in south-west Germany in summer 2007 both in time and space to error estimation of the data and tailored the observations to model output (e.g.,p y pprovide information of all kinds of precipitation types and to relate the COPS results

p ( gCOSMO-DE COSMO-EU forecasts By launching a web site an easy access toprovide information of all kinds of precipitation types and to relate the COPS results

to a broader perspective (longer time series and larger spatial domain) The durationCOSMO DE, COSMO EU forecasts. By launching a web site an easy access todata quicklooks and first order analysis has been enabled To make the datato a broader perspective (longer time series and larger spatial domain). The duration

f ill th ibilit t t ti ti ll h d l bl ddata, quicklooks and first order analysis has been enabled. To make the data

il bl t th i tifi it th d t hi d t th W ld D tof one year will open up the possibility to statistically approach model problems and available to the scientific community the data were archived at the World Databetter pin down specific model weaknesses. The GOP collected as many data about Center for Climate in Hamburg. By developing techniques to bring togetherbetter pin down specific model weaknesses. The GOP collected as many data aboutthe atmospheric state as possible within an area covering Germany and it

Center for Climate in Hamburg. By developing techniques to bring togetherobservations and model forecasts an optimal evaluation environment has beenthe atmospheric state as possible within an area covering Germany and it

i hb i t t F th th GOP ti i d th l it ti f i tiobservations and model forecasts an optimal evaluation environment has been

t dneighboring states. Furthermore the GOP optimized the exploitation of existing created.

D t S tData SetsData Sets

W th R d (GOP 2)R i G (GOP 1) DSD (GOP 3) GPS (GOP 5) S t llit (GOP 7)Weather Radar (GOP-2)Rain Gauges (GOP-1) DSD (GOP-3) GPS (GOP-5) Satellite (GOP-7)( )g ( ) ( ) ( ) ( )DWD rain gauge network: • Micro rain radar GPS network for integrated MSG:DWD rain gauge network:- daily sums: ~3000

Micro rain radar• Parsival

GPS network for integrated water vapor measurements • cloud mask- daily sums: ~3000

1 h l ti 700 DWD network: 16 C-Band• Parsival

J /W ld lwater vapor measurements • cloud mask

clo d top press re- 1 h resolution: ~700 DWD network: 16 C BandP l i t i h d

• Joss/Waldvogel • cloud top pressure - 1 min resolution: ~60 Polarimetric research radars • Optical disdrometers Lightning (GOP 6) • optical depth DWD 6-h analyses: - POLDIRAD (DLR)

p Lightning (GOP-6) p p• IR brightness temperatureDWD 6 h analyses:

RANIE gauge and radar( )

- Hohenpeißenberg (DWD)IR brightness temperature

MODIS:- RANIE gauge and radar Hohenpeißenberg (DWD)X B d d (UB ) Lightning for cloud-to-ground MODIS:

l d k- RANIE gauge only X-Band radar (UBonn) Lidar (GOP-4)Lightning for cloud to ground and in cloud discharges • cloud maskg g y

C-Band radar (FZ Karlsruhe)gaugel

gauge&

Lidar (GOP 4) and in cloud discharges LINET d SAFIR • cloud optical thickness τC-Band radar (FZ Karlsruhe)only &

radar • Earlinet (Hamburg,Leipzig, LINET and SAFIR cloud optical thickness τ• liquid water path LWPRMI radar Wideumont

Earlinet (Hamburg,Leipzig, Garmisch München)

• liquid water path LWPff ti di

DWD international compositGarmisch, München)DWD il t t k MetStat (GOP-8) • effective radius reff DWD international composit •DWD ceilometer network MetStat (GOP-8) • geometric cloud thickness H

DWD national composit ( > 100 stations in Germany) Cl d t t tig

• IWVDWD 24 h analysis:

pDWD online calibrated radar

( y)•several ceilometer stations • Cloudnet stations IWV

MERIS:DWD 24-h analysis: DWD online calibrated radar several ceilometer stations operated by universities/ • Meteorological stations MERIS:

- REGNIE precipitation (RADOLAN) operated by universities/ g• ARM mobile facility • cloud mask

• Berlin high density rainp p ( )

research institutes ARM mobile facilityCOPS stations • cloud optical thickness τBerlin high density rain

gauge network• COPS stations cloud optical thickness τ

• cloud top pressuregauge network • cloud top pressure

M d lEvaluation ModelsEvaluation Modelsa uat o

Emphasis on DWD‘s COSMO-EU (boundary) Bias and rmse in cloud base p ( y)and the convection-resolving COSMO-DE:height for model forecasts (red - and the convection-resolving COSMO-DE:

Grid size: Δx 2 8 km21h, green -12h, blue -3h)

• Grid size: Δx = 2.8 km (COSMO-DE left, COSMO-EU - Timestep: 25 s right) to ceilometer p

• 421 x 461 x 50 gridpoints - Lowest level 10measurements for all stations 421 x 461 x 50 gridpoints Lowest level 10 m abovewithin the model domains over m above

fthe whole GOP year 2007.

surfacey

• Centre of model domain:Bi d i i t t d Centre of model domain:10 °E 50 °N

Bias and rmse in integratedt f d l f t 10 E, 50 N

F t ti 21 hwater vapor for model forecasts(COSMO DE l ft COSMO EU • Forecast time: 21 h(COSMO-DE left, COSMO-EU i ht) t t f • Started every 3 hright) to gps measurements forll t ti ithi th d l

yall stations within the model d i th h l GOPdomains over the whole GOP

2007Up to 7 forecasts at a given time!

year 2007.Up to 7 forecasts at a given time!

D 1 Day 1

Day 1 Day 1 6 UTC

y9 UTC

y3 UTC

L d f t blLagged forecast ensemble

Day 1 00UTC

Day 2 00UTC+ 21 h forecast00UTC 00UTC 21 h forecast

C l iConclusionsConclusionsThe GOP gathered a large data set of in-situ and remote sensing observations for Central Europe with focus onThe GOP gathered a large data set of in situ and remote sensing observations for Central Europe with focus on water cycle A near realtime model evaluation for the COSMO EU and COSMO DE has been implementedwater cycle. A near realtime model evaluation for the COSMO-EU and COSMO-DE has been implemented. First analysis revealed:F t d d d l d b h i ht (l ft) d M diff b t i ht First analysis revealed:Forecasted and measured cloud base height (left) and

i t t d t ( i ht) f th diff t f t lMean difference between night-

ti d d ti IWV bi • both models overestimate precipitation by 20% (not shown) during winter timesintergrated water vapor (right) for the different forecast leas ti i di t d b th l

time and day-time IWV bias( di d ) both models overestimate precipitation by 20% (not shown) during winter timestimes indicated by the colors. (radiosonde-gps).

• integrated water vapor shows a significant diurnal cycle related to the daytime dry bias of radiosondesBIAS

0 UTC "specific humidity" soundings: 411

RMSE number of data/level • modeled cloud base height is significantly higher during day times due to assimilated dry biased radiosondesBias and RMSE10000BIAS

−0 h−3 h

RMSE10000

number of data/level • modeled cloud base height is significantly higher during day times due to assimilated dry biased radiosondesBias and RMSE in mixing ratio at9000

−3 h−6 h−9 h

9000 • high Bias in cloud base at North Sea stations indicate problem with different boundary layer abovein mixing ratio at

00UTC for all8000−12 h−15 h−18 h

8000

high Bias in cloud base at North Sea stations indicate problem with different boundary layer above00UTC for all radiosondes

7000

−18 h−21 h 7000 • data status and near-realtime evaluation results can be found at http://gop.meteo.uni-koeln.deradiosondes

within the6000

[m]

[m]

6000

[m] Outlook: The GOP is an ongoing project Current investigation focus on evaluation of the models

within theCOSMO DE

5000

eigh

t [m

eigh

t [m

5000

eigh

t [m Outlook: The GOP is an ongoing project. Current investigation focus on evaluation of the models

f ith t llit d t lik MSG AMSUCOSMO—DE

domain für J l4000

heig

heig

4000

heig performance with satellite products like MSG or AMSU.domain für July

2007 C l3000 3000

2007. Colors i di t diff t

2000 2000indicate different

f t l dBi d RMSE i l d b h i ht1000

2000

1000

2000

forecast leadti Crewell S M Mech T Reinhardt C Selbach H -D Betz E Brocard G Dick E O’Connor JBias and RMSE in cloud base height

d i d f il t t0

1000

0

1000 times. Crewell, S., M. Mech, T. Reinhardt, C. Selbach, H. D. Betz, E. Brocard, G. Dick, E. O Connor, J. Fischer T Hanisch T Hauf A Hünerbein L Delobbe A Mathes G Peters H Wernli M Wiegner

derived from ceilometer meauseumentsd COSMO DE d l f −0.5 0 0.5

0

specific humidity [g/kg]

0 0.5 1 1.5

specific humidity [g/kg]0 200 400

0 Fischer, T. Hanisch, T. Hauf, A. Hünerbein, L. Delobbe, A. Mathes, G. Peters, H. Wernli, M. Wiegnerand V Wulfmeyer 2008: General Observation Period 2007: Concept and first results Met Z accepted

and COSMO-DE model forecasts.and V. Wulfmeyer, 2008: General Observation Period 2007: Concept and first results. Met. Z. accepted