03/22/22 - 1 - COSMO Workpackage No. 5.1.8 First Results on Verification of LMK Test Runs Basing on SYNOP Data Lenz, Claus-Jürgen; Damrath, Ulrich [email protected]; [email protected]
Dec 30, 2015
04/19/23 - 1 -COSMO Workpackage No. 5.1.8
First Results on Verification of LMK Test Runs
Basing on SYNOP Data
Lenz, Claus-Jürgen; Damrath, [email protected]; [email protected]
04/19/23 - 2 -COSMO Workpackage No. 5.1.8
Overview
• survey on LMK test runs performed
• verification results
• summary and outlook
04/19/23 - 3 -COSMO Workpackage No. 5.1.8
Survey on LMK experiments performed (1)
- center of model area 10° E, 50° N
- 421 x 461 grid points horizontally
- grid length: 2.8 km
- 50 vertical layers, height of lowest layer: 40 m, height of lowest half level: 22 m above ground
- values for initialisation and boundaries from operational LM run
04/19/23 - 4 -COSMO Workpackage No. 5.1.8
Survey on LMK experiments performed (2)
testsuites performed for winter (Januar 2004)
experiment time integrationscheme, timestep
horizontaladvectionscheme
calculation ofprecipitation
thermo-dynamics
relaxation at lateralboundaries
673 leapfrogΔt = 16 s
centered2. order
diagnostic advection ofT=T0+T*
tanh function, implicit,similar to LM
683 TVD-RK 3. orderΔt = 30 s
upwind5. order
prognostic advection ofT=T0+T*
tanh function, as in LM,precipitation diagnostic
687 TVD-RK 3. orderΔt = 30 s
upwind5. order
prognostic advection ofT*=T-T0
cos function, explicit,precipitation diagnostic
688 TVD-RK 3. orderΔt = 30 s
upwind5. order
prognostic advection ofT*=T-T0
cos function, explicit,precipitation diagnostic,relaxation von qv, qcund qi
04/19/23 - 5 -COSMO Workpackage No. 5.1.8
Survey on LMK experiments performed (3)
testsuites for summer (July 2004) und autumn (September 2004)
- time integration: TVD Runge-Kutta 3. order, time step: Δt = 30 s- scheme for horizontal advection: upwind 5. order- prognostic calculation of precipitation- thermodynamics: advection of T=T0+T*- relaxation at lateral boundaries: explicit formulated cosine function
exp. parameterisationof shallowconvection
lateral boundaryvalues for pressure
precipitation scheme,number of classes
dataassimilation
advectionof qx
689 no interpolation 5 classes no Euler-forw.
696 yes vertical integration 5 classes no Euler-forw.
698 yes vertical integration including graupel6 classes
no Euler-forward
701 yes vertical integration 5 classes yes Euler-forw.
709 yes vertical integration including graupel6 classes
yes Semi-Lagrange
713 yes vertical integration including graupel6 classes
yes Bott-2-scheme
04/19/23 - 6 -COSMO Workpackage No. 5.1.8
LM and LMK:
TSS precipitation rates
> 0.1 mm/h (top left),
> 2 mm/h (top right),
> 10 mm/h (left)
January 2004, 00 h runs
04/19/23 - 7 -COSMO Workpackage No. 5.1.8
Frequency Bias (FBI) for
coverage due to low clouds
July 2004, 00 h runs
cloud coverage <= 2/8
cloud coverage >= 6/8
04/19/23 - 8 -COSMO Workpackage No. 5.1.8
Bias (mean error) for 2 m temperature, July 2004, 00 h runs
04/19/23 - 9 -COSMO Workpackage No. 5.1.8
LM and LMK:
TSS for precipitation rates
> 0.1 mm/h (top left),
> 2 mm/h (top right),
> 10 mm/h (left)
September 2004, 00 h runs
04/19/23 - 10 -COSMO Workpackage No. 5.1.8
Bias (mean error) for 2 m temperature, September 2004, 00 h runs
04/19/23 - 11 -COSMO Workpackage No. 5.1.8
sensitivity of TSS for precipitation
with respect to Latent Heat Nudging
07 - 19 July 2004, 12 h runs
prec. > 0.1 mm/h prec. > 2 mm/h
prec. > 10 mm/h
04/19/23 - 12 -COSMO Workpackage No. 5.1.8
sensitivity of TSS for precipitation with respect to interpolation method
LMK, experiment 701
September 2004, 12 h runs
prec. > 0.1 mm/h prec. > 2 mm/h
prec. > 10 mm/h
04/19/23 - 13 -COSMO Workpackage No. 5.1.8
Summary:
- gusts (not shown here): enhanced forecast accuracy in LMK due to finer grid resolution
- precipitation: results of the smaller scale model lead to partly better verification results than the LM routine forecast
- different tests (e.g. introduction of a parameterization scheme for shallow convection, data assimilation) lead to higher accuracy for some parameters, but other parameters may be affected negatively
- interpolation method during verification: higher TSS values at small precipitation rates and lower TSS values at higher precipitation rates using a larger averaging area than the nearest grid point
04/19/23 - 14 -COSMO Workpackage No. 5.1.8
Outlook:
- further continuous work on verification of past and future LMK experiments
- development of a tool for application of conditional verification is ongoing
- calculation of synthetic radar images using the radar simulation model by Haase (1998)
- verification of LM and LMK using the synthetic radar images