Name: JMA Ensemble
The S2S database contains real-time forecasts from JMA from 1st January 2015, and the full re-forecast dataset.
1. Ensemble version
Ensemble identifier code: GSM1403C
Short Description: Global ensemble system that simulates initial uncertainties using the bred vectors and lagged averaging forecasts and model uncertainties due to physical parameterizations using a stochastic scheme. Ensembles are based on 50 members: run once a week (Tuesday, Wednesday at 12Z) up to day 34.
Research or operational: Operational
Data time of first forecast run: 05/03/2014
2. Configuration of the EPS
Is the model coupled to an ocean model? No
If yes, please describe ocean model briefly including frequency of coupling and any ensemble perturbation applied: N/A
Is the model coupled to a sea Ice model? No
If yes, please describe sea-ice model briefly including any ensemble perturbation applied: -
Is the model coupled to a wave model? N/A
If yes, please describe wave model briefly including any ensemble perturbation applied: N/A
Ocean model: N/A
Horizontal resolution of the atmospheric model: TL319 (about 55 km)
Number of model levels: 60
Top of model: 0.1 hPa
Type of model levels: sigma
Forecast length: 34 days (816 hours)
Run Frequency: once a week (combination of Tuesday 12Z and Wednesday 12Z)
Is there an unperturbed control forecast included?: Yes
Number of perturbed ensemble members: 48 (2 controls from each initial date)
Integration time step: 20 minutes
3. Initial conditions and perturbations
Data assimilation method for control analysis: 4D Var
Resolution of model used to generate Control Analysis: TL959L100
Ensemble initial perturbation strategy: Bred vectors (extratropics (NH) plus tropics) + Lagged Average Forecasting
Horizontal and vertical resolution of perturbations: TL319L60
Perturbations in +/- pairs: Yes
Initialization of land surface:
1. What is the land surface model (LSM) and version used in the forecast model, and what are the current/relevant references for the model? Are there any significant changes/deviations in the operational version of the LSM from the documentation of the LSM? The Simple Biosphere (SiB) developed by Sellers et al.(1986), Sato et al.(1989a) and Sato et al.(1989b) has been implemented for the land surface process in forecast model.
2. How is soil moisture initialized in the forecasts? (climatology / realistic / other)? Is there horizontal and/or vertical interpolation of initialization data onto the forecast model grid? If so, please give original data resolution(s). Does the LSM differentiate between liquid and ice content of the soil? If so, how are each initialized? If all model soil layers are not initialized in the same way or from the same source, please describe. Initial soil moisture data (which consists of three layers) is produced by the offline simulations of the land surface model. The land processes in this simulations are similar to the one set in forecast model, and no horizontal and vertical interpolations are introduced in this analysis. Soil ice is handled as soil water, and no soil ice is used as initial condition specifically.
3. How is snow initialized in the forecasts? (climatology / realistic / other)? Is there horizontal and/or vertical interpolation of data onto the forecast model grid? If so, please give original data resolution(s) Are snow mass, snow depth or both initialized? What about snow age, albedo, or other snow properties? Initial snow data is also produced by using the offline simulation of the land surface model, and no horizontal interpolation is introduced. In the offline simulation, snow depth data is updated once a day (00UTC) by the two-dimensional Optimal Interpolation using the SYNOP snow depth data. The first guess is calculated by snow depth data from the offline simulation and snow cover data estimated by satellite observation. Forecast model calculates the snow water equivalent, so snow depth data is converted to the snow water equivalent. Snow density is set as a function related to the snow water equivalent (Verseghy 1991). Snow albedo is set as a function of wavelength and snow temperature. The age of snow is not considered.
4. How is soil temperature initialized in the forecasts? (climatology / realistic / other) Is the soil temperature initialized consistently with soil moisture (frozen soil water where soil temperature ≤0°C) and snow cover (top layer soil temperature ≤0°C under snow)? Is there horizontal and/or vertical interpolation of data onto the forecast model grid? If so, please give original data resolution(s) If all model soil layers are not initialized in the same way or from the same source, please describe. Soil temperature is also initialized by the offline simulations of the land surface model. No horizontal and vertical interpolation is implemented. Note that soil layers are three for soil moisture, while it is only one layer for soil temperature. Snow cover is updated at each 00UTC based on the snow depth analysis. At the same time, soil temperature for all grids where snow exists is set as less than 0deg. Once the soil temperature becomes less than 0deg, soil water changes soil ice (No consideration about freeze latent heat). No soil water scatters or moves in the freezing soil.
5. How are time-varying vegetation properties represented in the LSM? Is phenology predicted by the LSM? If so, how is it initialized? If not, what is the source of vegetation parameters used by the LSM? Which time-varying vegetation parameters are specified (e.g., LAI, greenness, vegetation cover fraction) and how (e.g., near-real-time satellite observations? Mean annual cycle climatology? Monthly, weekly or other interval?) There is no consideration about time-varying of vegetation properties. The climatology for each vegetation based on Dorman and Sellers(1989) is set as outer parameters (Some parameters are monthly data, such as LAI, the ratio of green leaves, the ratio of vegetation)
6. What is the source of soil properties (texture, porosity, conductivity, etc.) used by the LSM? The source of soil properties is different depending on the property. For example, soil porosity is set as outer parameters for each type of vegetation, while soil heat conductivity is set as a function related to the porosity and soil moisture in the first soil layer. In addition, some parameters are not considered such as difference of soil texture.
7. If the initialization of the LSM for re-forecasts deviates from the procedure for forecasts, please describe the differences. There is no difference between re-forecast and operational forecast about the procedure for initialization of land surface. For the re-forecast, land analysis data of JRA-55 is utilized as the land initial data and it is derived from the offline system forced by JRA-55 atmospheric field. This system is similar to the operational system, but note that the atmospheric forcing for operational offline system is given from the operational Global Analysis.
4. Model Uncertainties perturbations:
Is model physics perturbed? If yes, briefly describe methods: Stochastic physics
Do all ensemble members use exactly the same model version? Same
Is model dynamics perturbed? No
Are the above model perturbations applied to the control forecast? No
5. Surface Boundary perturbations:
Perturbations to sea surface temperature? No
Perturbation to soil moisture? No
Perturbation to surface stress or roughness? No
Any other surface perturbation? No
Are the above surface perturbations applied to the Control forecast? No
Additional comments: None
6. Other details of the models:
Description of model grid: Linear grid
List of model levels in appropriate coordinates: http://jra.kishou.go.jp/JRA-55/document/JRA-55_handbook_TL319_v2_en.pdf
What kind of large scale dynamics is used? Spectral semi-lagrangian
What kind of boundary layer parameterization is used? Mellor and Yamada level 2.5
What kind of convective parameterization is used? Arakawa and Schubert (JMA 2013)
What kind of large-scale precipitation scheme is used?: Sundqvist (1978)
What cloud scheme is used?: Smith (1990), Kawai and Inoue (2006)
What kind of land-surface scheme is used? SiB (Sato et al. 1989)
How is radiation parametrized? Outline of the operational numerical weather prediction at the Japan Meteorological Agency
Other relevant details?7. Re-forecast Configuration
Number of years covered: 30 years (1981-2010)
Produced on the fly or fix re-forecasts?: fixed re-forecasts in advance
Frequency: The re-forecasts consists of a 5-member ensemble starting the 10th, 20th, the last dates of calendar months.
Ensemble size: 5 members
Initial conditions: JRA-55 (TL319L60) + JRA-55 land analysis (TL319)
Is the model physics and resolution the same as for the real-time forecasts: Yes
If not, what are the differences: N/A
Is the ensemble generation the same as for real-time forecasts? Yes, except for lagged average forecasting.
If not, what are the differences: N/A
Other relevant information:
The JMA re-forecasts dataset is a "fixed" dataset which means that the re-forecasts are produced once from a "frozen" version of the model and are used for a number of years to calibrate real-time forecast. The JMA re-forecasts consist of a 5-member ensemble running three times a month from 1981 to 2010. The start dates correspond to 1st / 11th and 21st of each month at 00Z minus 12 hours (28 February instead of 29 February). here is the complete list of re-forecast start dates: 10/20/31 January - 10/20/28 February - 10/20/31 March - 10/20/30 April - 10/20/31 May - 10/20/30 June - 10/20/31 July - 10/20/31 August - 10/20/30 September - 10/20/31 October - 10/20/30 November and 10/20/31 December 1981-2010
The S2S database contains the complete JMA re-forecast dataset.
As for the other models, JMA re-forecasts are archived in the S2S database with 2 date attributes:
- hdate which corresponds to the actual starting date of the re-forecast
- date which correspond tot he ModelVersionDate.Since the JMA re-forecasts are "fixed" re-forecasts this ModelVersiondate is the same for all the re-forecasts and equal to 20140304. This variable will change when a new version of the JMA model for extended-range forecasts will be implemented.
Dorman, J. L. and P. J. Sellers, 1989: A global climatology of albedo, roughness length and stomatal resistance for atmospheric general circulation models as represented by the simple biosphere model (SiB). J. Appl.Meteor., 28, 833–855.
Kobayashi, S., Y. Ota, Y. Harada, A. Ebita, M. Moriya,H. Onoda, K. Onogi, H. Kamahori, C. Kobayashi, H. Endo, K. Miyaoka, and K. Takahashi, 2015: The JRA-55 reanalysis: General specifications and basic characteristics. J. Meteor. Soc. Japan. 93, 5-48.
Sato, N., P. J. Sellers, D. A. Randall, E. K. Schneider, J. Shukla, J. L. Kinter III, Y-T Hou, and E. Albertazzi, 1989a: Effects of implementing the simple biosphere model (SiB) in a general circulation model. J. Atmos. Sci., 46, 2757–2782.
Sato, N., P. J. Sellers, D. A. Randall, E. K. Schneider, J. Shukla, J. L. Kinter III, Y-T Hou, and E. Albertazzi, 1989b: Implementing the simple biosphere model (SiB) in a general circulation model: Methodologies and results. NASA contractor Rep. 185509, NASA. 76pp.
Sellers, P. J., Y. Mintz, Y. C. Sud, and A. Dalcher, 1986: A simple biosphere model (SiB) for use within general circulation models. J. Atmos. Sci., 43, 505–531.
Verseghy, D., 1991: CLASS—A Canadian Land Surface Scheme for GCMs. I. Soil model, Int. J. Climate., 13, 111 – 133.
Comprehensive description of the model and One-month ensemble prediction system: