ALARO Physics Neva Pristov (LACE area leader for physics) ALARO - - PowerPoint PPT Presentation

Regional Cooperation for Limited Area Modeling in Central Europe

ALARO Physics

Neva Pristov (LACE area leader for physics)

ALADIN Forecasters meeting, Sep 2014

ALARO

One of the physical parameterization package inside

ALADIN/HARMONIE system

Characteristics:

multi-scale: parameterizations being as scale-independent as

possible and giving physically consistent results over a wide range of model resolutions (in particular 10 km to 2 km)

consistency of all formulation, in particular within the 3MT

framework for the macro physical parameterization of precipitations

a prognostic character of parameterizations, while they share

the same information, approaches and level of complexity.

code stability, numerical efficiency and modularization

ALARO applications running at grey zone scales

ALADIN Forecasters meeting, Sep 2014

9km 4km Old CV 3MT 6h precipitation amounts. 3MT shows a reasonable solution at both resolutions. Old convection shows a very bad pattern, especially at 4km mesh (‘grey-zone syndrome’).

ALADIN Forecasters meeting, Sep 2014

ALARO-0 status

In the operational use in ALADIN countries

at, be, cz, hr, hu, po, ro, sk, si, tr, (po)

model resolution between 8 km – 4 km, 2km

In EPS systems

ALADIN-LAEF, GLAMEPS, EPS at HMS

In climatological simulations

be, cz, se

Plans for a usage in

HarmonEPS convection-permitting ensemble system multi model systems

ALADIN Forecasters meeting, Sep 2014

Versions:

ALARO-0-without3MT (January 2007) ALARO-0-with3MT (June 2008) ALARO-0 baseline (December 2012) ALARO-1 version1 (in preparation)

Workshops

Training Course ALARO-0, Mar 2007, Radostovice ALARO-1 Working Days, Feb 2010, Budapest ALARO-1 Working Days, Jun 2012, Ljubljana ALARO-1 Working Days, May 2014, Vienna

Short history

ALADIN Forecasters meeting, Sep 2014

ALARO-0

governing equations for moist processes in barycentric system radiation scheme NER (The Net Exchange Rate formulation) turbulence: pseudo-prognostic TKE moist physics:

Prognostic cloud water, cloud ice, rain and snow 3MT(Modular, Multiscale, Microphysics and Transport)

cloud and precipitation parameterization with prognostic microphysics unified for cloud water from resolved and unresolved processes; from the various condensation‐evaporation processes, sandwiched between updraft and downdraft

autoconversion, collection, evaporation/melting, freezing,.. statistical approach for sedimentation of rain and snow

Horizontal diffusion: SLHD

ALADIN Forecasters meeting, Sep 2014

ALARO-0

Organization of the physics time step

2 types of condensation 2 types of evaporation 1 radiation computation 1 microphysics computation

ALADIN Forecasters meeting, Sep 2014

Improvements and adaptations of parameterizations for

higher resolutions

Challenges:

Cost of radiation scheme Complexity of turbulence, still with a need to “parameterize” Advances in clouds and precipitation microphysics Keeping complementary parameterization of precipitating

convection

ALARO-1

ALADIN Forecasters meeting, Sep 2014

ALARO-1 developments (<10 km, down to 1 km)

Turbulence and shallow convection scheme TOUCANS Radiation ACRANEB2 Convection

Unsaturated downdraft scheme, CSD (Complementary Subgrid Drafts):

parameterization with a set of high resolution-specific features ascent, closure, triggering, evolution, deep convection

Cloud scheme

unified cloud treatment in radiation, shallow convection,

thermodynamic adjustment and 3MT,

Microphysics, prognostic graupel

Rain drop size distribution Geometry of cloud and precipitation

ALADIN Forecasters meeting, Sep 2014

Impact on diurnal cycle

average of mean hourly precipitation over the Czech area (11 realizations, Jun/Jul 2009) red - observations green - ALARO-0 “old” blue

• ALARO-0 baseline

Model starts rain early; Early decay as well; Too much precipitation in the morning; Lack of precipitation in the late afternoon and in the evening.

ALADIN Forecasters meeting, Sep 2014

Impact on diurnal cycle

average of mean hourly precipitation over the Czech area (11 realizations, Jun/Jul 2009) red - observations green - ALARO-0 “old” blue

• ALARO-0 baseline

magenta – ALARO-1step1 diurnal cycle of convection is improved

ALADIN Forecasters meeting, Sep 2014

SAL verification

Czech area 22 Jun – 7 Jul 2009

0.5mm/3hours 10mm/3hours

Structure component: horizontal axis Amplitude component: vertical axis Location: color (red – best L<0 0-1, green, blue, black for L>1)

ALADIN Forecasters meeting, Sep 2014

ALARO-0 baseline version

grey-zone experiment defined by WGNE group

(http://www.knmi.nl/samenw/greyzone/index.html)

cold-air outbreak case

convection develops over warmer sea surface no big problem with orography

simulated at various resolutions:

16km, 8km, 4km, 2km and 1km over north Atlantic

run from 30 January 2010, 12 UTC, up to 36h without and with parameterized moist deep convection a very tough test-bed for multi-scale convection

parameterizations alike 3MT

ALADIN Forecasters meeting, Sep 2014 16km 8km 4km 2km 1km MODIS radiance

• bservatio

n

WGNE grey-zone test, ALARO-0, cloud cover at 24h range

ALADIN Forecasters meeting, Sep 2014 16 km 8 km 4 km 2 km 1 km

WGNE grey-zone test, ALARO-0,

1h precipitation cumulated (30.1.2010 1231h), area between Faeroe and Orkney islands. The general structure of this short time precipitation pattern is kept, while more detailed solutions continuously appear with finer model resolutions.

ALADIN Forecasters meeting, Sep 2014 16 km 8 km 4 km 2 km 1 km

WGNE grey-zone test, ALARO-0, 1h precipitation (30.1.2010 12+31h)

3MT activated 3MT disabled

ALADIN Forecasters meeting, Sep 2014

2 options for mixing length computation

14.12.2012 00:00 +34h Wind at 10 m pTKE AY pTKE EL1 Default z-dependent mixing length Mixing length dependent on TKE False alarm for SW wind

ALADIN Forecasters meeting, Sep 2014

2 options for mixing length computation

14.12.2012 00:00 +34h Wind at 10 m pTKE AY pTKE EL1 False alarm for SW wind

ALADIN Forecasters meeting, Sep 2014

2 options for mixing length computation

Vertical cross-section Temperature TKE pTKE AY pTKE EL1 Default z-dependent mixing length Mixing length dependent on TKE

ALADIN Forecasters meeting, Sep 2014

Convection diagnostics

mixed layer CAPE storm motion vector, vertical wind shear, relative

helicity

lightning diagnostics

4 different methods in test

diagnosed hail

vertical integral of graupel, as an instantaneous flux maximum over a given period

ALADIN Forecasters meeting, Sep 2014

Type of the process Location and timing can differ from reality

Simulated radar reflectivity Lightning density

ALADIN Forecasters meeting, Sep 2014

Severe freezing rain case

• end of January 2014

ALADIN Forecasters meeting, Sep 2014

Severe freezing rain case

ALADIN Forecasters meeting, Sep 2014

Severe freezing rain case - analysis

Temperature BIAS Strength of inversion is diminishing during forecast

ALADIN Forecasters meeting, Sep 2014

Severe freezing rain case

ALADIN Forecasters meeting, Sep 2014

www.rclace.eu

Operational products

ALADIN Forecasters meeting, Sep 2014

Summary

ALARO parameterizations

past and future developments testing, validation examples multi-scale performance

Some strengths and weaknesses

realistic convection features diurnal cycle of convective precipitation, light precipitation, melting layer

Interpretation of the model results

(new) model – higher resolution – new challenges some processed are more realistically described Is the forecasting process able to cope with these changes?