1. Forecast system version
Identifier code: JMA/MRI-CPS4
First operational forecast run: January 2026
2. Configuration of the forecast model
Is the model coupled to an ocean model? Yes
Coupling frequency: 1 hour
2.1 Atmosphere and land surface
| Model | JMA-GSM |
|---|---|
| Horizontal resolution and grid | TL319 (approx. 55km) |
| Atmosphere vertical resolution | L128 |
| Top of atmosphere | 0.01hPa |
| Soil levels (layers) | 7 Layer 1 : 0 - 0.02 m |
| Time step | 20 minutes |
Detailed documentation:
JMA, 2019: Outline of the operational numerical weather prediction at the Japan Meteorological Agency. Appendix to WMO Technical Progress Report on the Global Data-processing and Forecasting System and Numerical Weather Prediction Research. Appendix to WMO Technical Progress Report on the Global Data-processing and Forecasting System and Numerical Weather Prediction, Tokyo, Japan.
2.2 Ocean and cryosphere
| Ocean model | MRI.COM v5.0 |
|---|---|
| Horizontal resolution | 0.25ºx0.25º on a tripolar grid |
| Vertical resolution | L60 |
| Time step | 10 minutes |
| Sea ice model | part of MRI.COM v5.0 |
| Sea ice model resolution | same as ocean model |
| Sea ice model levels | 5 categories + open water |
| Wave model | None |
| Wave model resolution | N/A |
Detailed documentation:
3. Initialization and initial condition (IC) perturbations
3.1 Atmosphere and land
| Hindcast | Forecast | |
|---|---|---|
| Atmosphere initialization | Japanese Reanalysis for Three Quarters of a Century (JRA-3Q) | Global Analysis (GA) |
| Atmosphere IC perturbations | Singular Vectors (SV) | Singular Vectors (SV) and Local Ensemble Transform Kalman Filter (LETKF) |
Land Initialization | Offline model runs forced by JRA-3Q | Offline model runs forced by JRA-3Q and GA |
| Land IC perturbations | None | None |
| Soil moisture initialization | Offline model runs forced by JRA-3Q | Offline model runs forced by JRA-3Q and GA |
| Snow initialization | Offline model runs forced by JRA-3Q | Offline model runs forced by GA |
| Unperturbed control forecast? | Yes | Yes |
Data assimilation method for control analysis: 4D-VAR
Horizontal and vertical resolution of perturbations: TL319L128 (LETKF), TL63L40 (SV)
Perturbations in +/- pairs: Yes
3.2 Ocean and cryosphere
| Hindcast | Forecast | |
|---|---|---|
| Ocean initialization | Multivariate Ocean Variational Estimation (MOVE)/MRI.COM Global 3 system (MOVE/MRI.COM-G3) | Multivariate Ocean Variational Estimation (MOVE)/MRI.COM Global 3 system (MOVE/MRI.COM-G3) |
| Ocean IC perturbations | Ensemble perturbations approximating analysis error covariances using minimization histories | Ensemble perturbations approximating analysis error covariances using minimization histories |
| Unperturbed control forecast? | Yes | Yes |
Detailed documentation:
https://www.data.jma.go.jp/wmc/products/elnino/move_mricom-g3_doc.html
4. Model Uncertainties perturbations:
| Model dynamics perturbations | None |
|---|---|
| Model physics perturbations | Stochastically Perturbed Parametrization Tendencies (SPPT) scheme and Stochastic Humidity Profile for Convective parameterization (SHPC) |
If there is a control forecast, is it perturbed? | No |
Detailed documentation:
5. Forecast system and hindcasts
| Forecast frequency | daily |
|---|---|
| Forecast ensemble size | 5 per day |
| Hindcast years | 1991-2020 |
| Hindcast ensemble size | 10 (5 members with 15-day Lagged Average Forecast) |
| On-the-fly or static hindcast set? | static |
6. Other relevant information
The available hindcast start dates are as follows:
| Start Month | Available Start Days |
|---|---|
| January | 16 and 31 |
| February | 10 and 25 |
| March | 12 and 27 |
| April | 11 and 26 |
| May | 16 and 31 |
| June | 15 and 30 |
| July | 15 and 30 |
| August | 14 and 29 |
| September | 13 and 28 |
| October | 13 and 28 |
| November | 12 and 27 |
| December | 12 and 27 |
7. Where to find more information
General information:
https://www.data.jma.go.jp/wmc/products/model/outline/cps_description.html
https://www.data.jma.go.jp/wmc/products/elnino/move_mricom-g3_doc.html
