first results of the cosmo priority project ’tackle deficiencies in … · 2016-12-20 · pp qpf...

Federal Department of Home Affairs FDHAFederal Office of Meteorology and Climatology MeteoSwiss

First results of the COSMO Priority Project ’Tackle deficiencies in

quantitative precipitation forecasts’

LM User Seminar, 6 March 2007, Langen

S. Dierer1, J. Achimowicz9, T. Andreadis8, M. Arpagaus1, E. Avgoustoglou8, M. Baldauf2, C. Cacciamani3, U. Damrath2, R.

Dumitrache10, V. Fragkouli8, A. Fuccello4, F. Grazzini3, T. La Rocca4, P. Louka8, P. Mercogliano7, P. Mezzasalma3, M. Milelli5, E. Oberto5, A. Parodi6, I. V. Pescaru10, U. Pflüger2, P. Prohl2, I. Pytharoulis8, A.

Sanna5, Ch. Schraff2, F. Schubiger1, A. Seifert2, K. Starosta9, J. Steppeler2, M. Tesini3, E. Zala1

1MeteoSwiss (CH), 2DWD (D), 3ARPA-ER (IT), 4UGM (IT), 5ARPA-P (IT), 6Uni Genova (IT), 7CIRA (IT), 8HNMS (GR), 9IMGW (PO), 10NMA (RO)

2PP QPF

S. Dierer et al.

Outline

• Aim of COSMO priority project ’Tackle deficiencies in quantitative precipitation forecasts’ (PP QPF)

• Overview of PP QPF• Selection of test cases• List of sensitivity studies • First results of the sensitivity studies• Conclusions and Outlook

3PP QPF

S. Dierer et al.

Aim of PP QPF

The aim of PP QPF is improved knowledge about • most suitable namelist settings or • parts of the model that need to be reformulated

to obtain a better QPF at 7 km horizontal grid size

Good quantitative precipitation forecast is a challenging task – not only but also in the COSMO model:

The project has a focus on model deficiencies – not on errors from e.g. large scale conditions

4PP QPF

S. Dierer et al.

Overview of PP QPF

• Task 1: Selection of test cases representative for„typical“ QPF deficiencies of COSMO model

• Task 2: Definition of sensitivity studies

• Task 3: Run sensitivity studies and draw conclusions

• Task 4: Idealized runs for moist benchmark cases

5PP QPF

S. Dierer et al.

Selection of test cases

6PP QPF

S. Dierer et al.

Selection of cases of poor QPF

• Selection of single cases reflecting general problems of QPF in every country

• Large scale situation should be well predicted in order to concentrate on poor performance of COSMO model

• Concentration on “newer” cases (2005 or later) in order to avoid deficiencies due to an old model version

• To ensure this, cases are rerun with reference version• Every country defines of final list of 3 test cases

7PP QPF

S. Dierer et al.

Example: 02.02.2005 - test case selected by MeteoSwiss

24h precipitation of all situations in 2005 withnortherly large scale flow: average precipitation patternfrom radar and forecast

24h precipitation on 02.02.2005: radar and forecast(northerly large scale flow)

8PP QPF

S. Dierer et al.

List of test cases from all countries

DATE INSITUTION Overestimation(+)/underestimation (-)

Stratiform(strat)/convective(con)

06.12.2004 LM-DWD + strat warm sector

18.03.2005 LM-DWD + strat orography

02.07.2005 LM-NMA +/- strat Cold front

03.05.2005 LM-DWD + Strat+conv Warm front

21.06.2005 LM-DWD - conv Cold front

02.02.2005 aLMo + strat Cold front

22.03.2005 aLMo + strat Warm front

12.07.2005 aLMo + conv -

17.12.2005 aLMo + Mainly strat orography

24.09.2004 LAMI - conv Cold front

10.04.2005 LAMI + conv Occluded front

17.08.2006 LAMI +/- conv front

09.09.2005 LAMI + conv -

27.03.2005 LAMI - strat/conv front

01.12.2005 EuroLM - mainly strat ?

03.12.2005 EuroLM +/- Strat/conv ?

17.12.2005 EuroLM - strat/conv ?

15.09.2005 LM-HNMS + conv orography

23.11.2005 LM-HNMS + Strat+conv Cold front

26.11.2005 LM-HNMS +/- Strat+conv Warm+cold front

03.05.2005 LM-IMGW + strat Warm front

04.05.2005 LM-IMGW +/- conv cold front

09.06.2005 LM-IMGW - conv occluded

09.08.2005 LM-IMGW +/- ? cold front

23.06.2005 LM-NMA + strat+conv cold front

12.07.2005 LM-NMA + conv -

9PP QPF

S. Dierer et al.

Frequency of types with poor QPF‘s

underestimation Mixed(under-/overestimation)

overestimation

Preliminary list 11 (20%) 13 (24%) 30 (56%)Final list 6 (30%) 5 (25%) 9 (45%)

by Ulrich Damrath, DWD

10PP QPF

S. Dierer et al.

Weather situations associated withpoor QPF

Preliminary list Final listWarmfront 8 (15%) 3 (15%)

Mixed (WF/CF) 3 (6%) 1 (5%)Coldfront 22 (41%) 9 (45%)Occlusion 11 (20%) 3 (15%)

Trough 5 (9%) 2 (10%)No front 5 (9%) 2 (10%)

by Ulrich Damrath, DWD

11PP QPF

S. Dierer et al.

List of sensitivity runs

12PP QPF

S. Dierer et al.

Different fields of sensitivity studies

• 1. Changes of initial conditions• 2. Changes of numerical methods• 3.1 Changes of microphysics• 3.2 Changes of convection schemes• 3.3 Changes of PBL schemes

13PP QPF

S. Dierer et al.

List of sensitivity studies

by Axel Seifert, DWD

14PP QPF

S. Dierer et al.



15PP QPF

S. Dierer et al.



16PP QPF

S. Dierer et al.



17PP QPF

S. Dierer et al.



18PP QPF

S. Dierer et al.

Runs of sensitivity runs

19PP QPF

S. Dierer et al.

0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

06.12.2004

23.06.2005

02.07.2005

18.03.2005

03.05.2005

21.06.2005

02.02.2005

22.03.2005

12.07.2005

17.12.2005

24.09.2004

10.04.2005

17.08.2006

09.09.2005

27.03.2005

01.12.2005

03.12.2005

17.12.2005

15.09.2005

23.11.2005

26.11.2005

03.05.2005

04.05.2005

10.06.2005

09.08.2005

12.07.2005

22

Runs of sensitivity studies (finished end 02/2007)

20PP QPF

S. Dierer et al.

First results from sensitivity studies for German, Swiss and Romanian cases

• All together 10 cases:• 5 cases of overestimated winter precipitation• 4 cases of overestimated summer precipitation• 1 case of underestimated summer precipitation

• Comparison of 24h area average precipitation of sensitivity study, control run and observations

21PP QPF

S. Dierer et al.

Winter cases with overestimated precipitation - change of 24h area average precipitation sum in the sensitivity studies

06.12.2004 18.03.2005

22PP QPF

S. Dierer et al.


02.02.2005 22.03.2005

5 dt206 lfsl37 Rksl38 rkbott9 rktp10 oro

11 micro112 micro213 micro314 conmod15 conkfb16 conoff

17 rlam0118 rlam5019 sto05020 sto25021 sea0122 sea40

O ctrl1 wso802 Wso1203 Qv0904 Qv110

23PP QPF

S. Dierer et al.


17.12.2005





24PP QPF

S. Dierer et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

06.12.200418.03.200502.02.200522.03.200517.12.2005

22






25PP QPF

S. Dierer et al.

Summer cases with overestimated precipitation - change of 24h area average precipitation sum in the sensitivity studies

03.05.200512.07.2005





26PP QPF

S. Dierer et al.


AVERAGE PRECIPITATION23.06.2005

Obse

rvat

ion

Cont

rol

WSO

080

WSO

120

QV 0

90

QV 1

10DT

20

LFS

13RK

S 13

RK b

ott

RK tp

Mic

ro 1

Mic

ro 2 M

icro

3CO

N m

od CON

off

RLAM

01

RLAM

50

STO

050

STO

250

0

2

4

6

8

10

12

14

(blue: grid-scale rain, red: convective precipitation)

mea

n ac

c.pr

ecip

in m

m

AVERAGE PRECIPITATION 02.07.2005

STO

250

STO

050

RLAM

50

RLAM

01

CON

off

CON

mod

Mic

ro 3

Mic

ro 2

Micr

o 1

RK tp

RK b

ott

RKS

13

LFS

13DT 20

QV 1

10

QV 09

0

WSO

120

WSO

080

cont

rol

obse

rvat

ion

0

5

10

15

20

25

(blue: grid-scale rain, red: convective precipitation)

mea

n ac

c. p

reci

p in

mm

27PP QPF

S. Dierer et al.

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21

03.05.200523.06.2005 ?02.07.2005 ?12.07.2005

22






28PP QPF

S. Dierer et al.

Summer cases with underestimated precipitation - change of 24h area average precipitation sum in the sensitivity studies

21.06.2005

29PP QPF

S. Dierer et al.

Conclusions and Outlook• None of the studies completely solves a QPF problem, but some

give a significant improvement for single cases like • changes of snow microphysics for winter precipitation

overestimation• Kain-Fritsch/Bechtold for summer precipitation underestimation

• Initial humidity is of great importance • Runge-Kutta reduces mean precipitation in most of the cases• Decreased vertical exchange of heat and moisture (RLAM) tends to

decrease mean precipitation• Next steps:

• Complete statistics for all runs• Detailed look at runs with changes in order to understand

changes, e.g. why does Runge-Kutta tend to decrease average precipitation? Why decreased vertical exchange?

• Possibly further studies, e.g. using COSMO model version 3.22

first results of the cosmo priority project ’tackle deficiencies in … · 2016-12-20 · pp qpf...

Documents