|
Russian activities in polar studies related to weather forecasting M. Tolstykh, Hydrometcentre of Russia
|
tarix | 23.01.2018 | ölçüsü | 445 b. | | #22289 |
|
M.Tolstykh, Hydrometcentre of Russia; Institute of Numerical Matematics Russian Academy of Sciences
Plan Background Observations Modelling
Many meteorological stations in Russian Arctica were closed/conserved in 1990s Many meteorological stations in Russian Arctica were closed/conserved in 1990s Today Arctic is getting one of Russian priorities Enhancement of forecasting capabilities in Arctic is essential for: - sea transport, - etc. International Polar Year 2007-2008 Polar forum in Moscow, September 2010 – Russian PM declared support for basic researches in Arctic Proposal for international polar decade
Organizations involved Roshydromet: observations Hydrometcentre of Russia, Arctic and Antarctic Research Institute: forecasts, diagnostic studies, modelling Institute of Numerical Mathematics RAS – ocean and ice modelling, coupled modelling (climate) Norway-Russian cooperation: StPetersburg NIERSC (mostly climate studies, but also some meteorology)
Russian activities in Antarctic 5 stations + 3 automatic weather stations Progress station is being built
Building the permanent complex on Progress station
Modernization of existing stations in Antarctida
Observations in Arctica: Current state and perspectives Density index for SYNOP stations 35-50 ( 1 station per 35000-50000 sq. km) in Russian Arctic as of 01.01.2009. It is planned to increase the number of SYNOP stations in Russia to 2200-2300 (among them 500-600 automatic), significant portion of this increase will be above 60N . Arctic stations will be mostly automatic). About 50 stations are being installed this year. Radiosondes: planned increase from 105 to 129 sites, soundings 2 times per day. Creation and maintenance of the “Arktika” satellite system including two Molniya satellites on high-elliptic orbits.
Drifting station “Severnii Polus 37 ” (09/2009-01/2010)
Short and medium-range numerical weather prediction Global models (SL-AV and T169L31) LAM – COSMO-RU 7 km Data assimilation Global ensemble prediction system (under testing) – based on T169 and SL-AV global models
SL-AV model Semi-Lagrangian vorticity-divergence dynamical core of own development, ALADIN/LACE parameterizations Currently, 0.9x0.72 degrees lon/lat, 28 levels, runs on Altix 4700 computer Accepted as the principal global model of RHMC Version with the resolution 0.45x0.37 degrees, 50 levels under testing Mire parametirization of own development Forecasts also at http://meteoinfo.ru/plav-forc-rus Spectral T169L31 model also runs operationally
T and RH errors at 2m for near-mire stations (bias, Abs and RMS errors)
Initial and boundary data: - Initial and boundary data:
- 00 and12 UTC, GME (DWD)
- Forecast: 78 h
- Grid step 7 km
- Grid:
- 700 * 620 * 40
- SGI Altix 4700
- (1664 cores)
- Run time for 78 h.
- 19 min: 1024 cores
-
Data assimilation 3D-Var-FGAT for global models under quasioperational tuning (M.Tsyrulnikov). LETKF for SL-AV global model is being implemented (A.Shlyaeva). Some form of hybrid is considered for far future
Domain: global Domain: global Perturbations of initial data: Breeding Regional rescaling (northern extratropics, southern extratropics, tropics) Members: 15 T85L31 – 12 perturbed + control + T169L31, unperturbed + SLAV, unperturbed Implementation: Quasi operational runs on Altix 4700.
Operational: 1)Seasonal version of SL-AV model Horizontal resolution 1,40625°х1,125° lon-lat, 28 vertical levels; 2) T42L14 (T63L25 soon) run by Main Geophysical Observatory (St.Petersburg) The models contribute to the multi-model ensemble of APCC. Forecasts are at http://wmc.meteoinfo.ru/season There is an experimental coupled atmosphere-ocean-sea ice model
Experimental atmosphere-ocean-ice model -1 Joint work of Hydrometcentre of Russia and Institute of Numerical Mathematics RAS Atmospheric part is SL-AV (1.4x1.1, 28 levels) Ocean and ice models, as well as the cooupler, are taken from the climate model of INM. Calibrated in CMIP4, participating in CMIP5.
sigma-coordinate model with isopicnic horizontal diffusion 1˚x0.5˚ , 40 levels The EVP (elastic- viscous- plastic) dynamics, Semtner thermodynamics sea ice model (Hunke, Ducowicz 1997; Iakovlev, 2005) is embedded. Coupling to the atmospheric model without flux correction
Experimental atmosphere-ocean-ice model -2 Averaged heat flux to the ocean is 2.8 W/m2 (evaluated on the ocean side while running the model with one-way interaction). Room for improvement! No flux correction. 10-member ensemble (only atmospheric initial data were perturbed) Requires 12 hours on 12 processors (one ensemble member).
Change of SST in one of the experiments: reanalysis (left), coupled model (right)
Some results According to first results, small positive changes in extratropics atmosphere as compared with extrapolated SST anomaly run. Recently, the results for SST forecasts were improved. Currently, hindcasts with the coupled model are being run.
But… Even potential predictability in Arctic on seasonal time scales is not too high… Plan to study if there is a positive impact of using coupled model for shorter time ranges (e.g. month)
Plans (related to the models) Increase of the resolution for medium range SL-AV model to 0.45x0.37 deg, 50 levels – parallel runs expected in 2011. COSMO-RU 7 km will have a special domain for Siberia (2011-2012) Coupled system – increase of the ocean model resolution to 0.25x0.25 deg, increase the number of vertical levels in atmospheric model from 28 to 40 (2012). 3D-Var-FGAT for the atmosphere and ocean expected operationally in 2011
Thank you for attention!
Dostları ilə paylaş: |
|
|