EXTPAR: Developments towards version 2.0

15
EXTPAR: Developments towards version 2.0 D. Lüthi, Institute for Atmospheric and Climate Science, ETH Zürich work contributed by: Martina Messmer, Anne Roches, Jürgen Helmert, Gerhard Smiatek, Jean-Marie Bettems, Frank Brenner, Eva-Maria Gerstner, Hermann Asensio, Günther Zängl COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 1

description

EXTPAR: Developments towards version 2.0. D. Lüthi, Institute for Atmospheric and Climate Science, ETH Zürich w ork contributed by: Martina Messmer , Anne Roches, Jürgen Helmert , Gerhard Smiatek, Jean-Marie Bettems , Frank Brenner, Eva-Maria Gerstner, Hermann Asensio , G ünther Zängl. - PowerPoint PPT Presentation

Transcript of EXTPAR: Developments towards version 2.0

Page 1: EXTPAR: Developments towards version 2.0

EXTPAR: Developments towards version 2.0

D. Lüthi, Institute for Atmospheric and Climate Science, ETH Zürich

work contributed by:Martina Messmer, Anne Roches, Jürgen Helmert,

Gerhard Smiatek, Jean-Marie Bettems, Frank Brenner, Eva-Maria Gerstner, Hermann Asensio, Günther Zängl

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 1

Page 2: EXTPAR: Developments towards version 2.0

Status of EXTPAR

• On March 4th the COSMO STC declared EXTPAR as an official COSMO software

• EXTPAR has to follow in the future the same development process rules as COSMO or int2lm

• There needs to be a source code administrator for this tool• STC appointed Daniel Lüthi as SCA• Responsible SMC coordinator for EXTPAR is Jean-Marie

Bettems (WG3b)

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 2

Page 3: EXTPAR: Developments towards version 2.0

Why is CLM-Community interested in EXTPAR?

• Groups of CLM-Community perform climate simulations outside the European domains

• Increasing number of CLM-Community members run simulations at cloud resolving scales

• Strong need of high quality external data sets globally at high resolution

• Need for documentation of external data• Long term reproducibility of external data sets• EXTPAR can deliver all of these requirements!

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 3

Page 4: EXTPAR: Developments towards version 2.0

Latest versions of EXTPAR

• Latest common version EXTPAR: 1.6• Latest version of EXTPAR at DWD: 1.12• Latest version of EXTPAR at MeteoSwiss: 1.14• Latest CLM-community version of EXTPAR: 1.6_clm6• New version of EXTPAR under way: 2.0

– Includes all recent developments by DWD, MeteoSwiss and CLM-community

– Support all available features and options of COSMO 5.0

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 4

Page 5: EXTPAR: Developments towards version 2.0

5 Improved external Parameters for the COSMO Model | 21.8.2013Martina Messmer

• 2 km resolution of COSMO-2 is still too coarse for Swiss topography

• Simulations with COSMO-1 are required for more precise forecasts

• BUT: resolution of GLOBE is similar to COSMO-1, meaning that sub-grid scale topography is not visible

A new high resolution digital elevation model (DEM) is needed to derive topography related external parameters.

Both SRTM and ASTER are considered.SWISSTOPO data are used as reference over CH.

GLOBE SWISSTOPO ASTER GDEM2 SRTM V4

Resolution 30 arc-sec (~1km) 25 meters 1 arc-sec (~30m) 3 arc-sec (~90m)

Lat range 90° N – 90° S whole CH 83° N – 83° S 60° N – 58° S

Projection WGS84 CH-1903 WGS84 WGS84

Topography:Motivation

Page 6: EXTPAR: Developments towards version 2.0

6 Improved external Parameters for the COSMO Model | 21.8.2013Martina Messmer

COSMO1 oro-smooth ASTER [m] GLOBE [m] REAL [m]BLUE 7.08476° E, 46.1324° N 860.1746 1455.024 828.6WHITE 7.08530° E, 46.1224° N 488.1573 936.2092 460.0

A zoom into Martigny clarifies that the ASTER based topography is closer to reality than the GLOBE based.

Topography:ASTER versus GLOBE, geolocation

Page 7: EXTPAR: Developments towards version 2.0

Topography relatedAdaptations in EXTPAR• Support for ASTER data (but currently limited to 60°N – 60°S)• There is a new switch, which allows user to choose between

GLOBE and ASTER• Two new parameters have been introduced, which allow to use

a subsample of the full ASTER domain (performance)ASTER is only recommended for very high resolutionapplications and for domains between 60°N – 60°S

• Parameters needed for topographic shading (lradtopo = .TRUE.) in the radiation code can now be produced within EXTPAR (Anne R.). But then you must also use the topography smoothing in EXTPAR rather than in int2lm

• For subgrid scale orography parameterization (lsso = .TRUE.) scale separation can be chosen (only with GLOBE dataset). A filtered dataset has been prepared for that.

Page 8: EXTPAR: Developments towards version 2.0

Topography:Scale Separation

• Raw GLOBE topography filtered to remove scales < 3 km• Not possible to obtain an equal filter in both directions.

-10%

~ 4 km

Page 9: EXTPAR: Developments towards version 2.0

Improved Parameters: Land-use

• Old Globcover data exhibits a shift around 44 °N.• Shift is corrected in raw data set (new 6 tiles instead of 1)• Additional shift of half a grid point with Globcover (bug fix)

Land area fraction:

left:GLC2000 – GC old(1km) (300 m)

right:GLC2000 – GC new

COSMO-1 COSMO-1

Page 10: EXTPAR: Developments towards version 2.0

EXTPAR Developments at DWD

• New background albedo climatology (12 monthly values) (F. Brenner)– Based on MODIS satellite data– Available for visible, near-infrared and ultraviolet spectral bands

(currently only visible is used in COSMO)– Consistency check only with FAO soil types

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 10

JAN (FAO) [%] COSMO-1

Page 11: EXTPAR: Developments towards version 2.0

EXTPAR Developments at DWD• Harmonized world soil database (HWSD) is interpolated to

COSMO/(ICON) grid (J. Helmert)– DWD implementation needs new versions of int2lm and COSMO will

not yet be supported in COSMO 5.0– soil information is available for two layer depths 0-30cm and 30-

100cm– Resolution 30” (~1km)– New soil parameters

• OpenMP parallelization of certain parts of EXTPAR (G. Zängl)

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 11

Page 12: EXTPAR: Developments towards version 2.0

Soil compounds (HWSD)FR_SAND [%] FR_SILT [%]

FR_OC [%]

FR_CLAY [%]

SOILTYP COSMO-2

Page 13: EXTPAR: Developments towards version 2.0

EXTPAR Developments in CLM-community

• Harmonized world soil database (HWSD) is interpolated to COSMO/(ICON) grid (G. Smiatek)– Direct mapping to TERRA soil types– Can be used with current versions of int2lm and COSMO

• Soil albedo with MODIS based soil color data (D. Lüthi)– Soil albedo treatment similar to Community Land Model– Data set 0.5 deg resolution– Yields better 2m temperatures over Saharan desert

• ECOCLIMAP as alternative land use data set (G. Smiatek)– Lookup table for monthly values of LAI, PLCOV, ROOTDP

• Support for alternative aerosol climatologies– AEROCOM (Lüthi), MACC-II (Panitz, Kothe)

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 13

Page 14: EXTPAR: Developments towards version 2.0

Next Steps

• Complete merging of code contributed by CLM-community to version 1.14 (D. Lüthi)

• Add full support for GRIB output (J.-M. Bettems)• Complement technical documentation of EXTPAR• Release version 2.0• Build framework for technical testing• Clean up code and improve usability• Explore possibility of speeding up execution by

parallelization• Scientific evaluation of new external parameter fields

COSMO General Meeting 2013 D.Lüthi: EXTPAR developments towards version 2.0 14

Page 15: EXTPAR: Developments towards version 2.0

Thank you for your attention!!