This web page documents the current† usage of Radio Occultation data by operational NWP centers. Summary data is shown in the tables below but the links in the ‘More info.’ column provide further details, as provided by each center such as model resolution, observation errors, etc. These pages should also include results of forecast impact/sensitivity studies where available.
† Centers will be prompted to provide up-to-date information annually, though major operational changes can be documented here at any time.
Summary table:
| Institute | Assimilation system | RO operator details | More info. |
Info valid for: | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Global | LAM | BA/REF | Operator | Bias corr. | TPD | |||||||||
| Type | Window | Early cutoff from start | Late cutoff from start | Early cutoff from end | Late cutoff from end | Loss rate (*see below) | ||||||||
| CMA (China) |
3D-Var | 6h | 6h30-8h00 | 10h40-13h00 | 0h30-2h30 | 4h40-7h00 | 17%/hour | 3D-Var (6hr) | REF | 1D | No. | No. | here | 24 Nov 2015 |
| DWD (Germany) |
3D-Var | 3h | 3h44 | 12h41 | 0h44 | 9h41 | 33%/hour | N/A | BA | 1D | No | No | here | 12 Nov. 2015 |
| Météo France (France) |
4D-Var | 6h | 4h50-6h00 | 9h55-11h05 | -1h10-0h00 | 3h55-5h05 | 17%/hour | 3DVar (6hr) | BA | 1D | No | Yes | here | 31 Oct. 2014 |
| Environment Canada (Canda) |
EnVar/ EnKF | 6h | 5h30 | 9h-12h | -0h30 | 3h00-6h00 | 17%/hour | 3DVar (4.7Hr) | REF | 1D | No | No | here | 2 Nov. 2015 |
| ECMWF (Europe) |
4D-Var | 6h/12h (early/late) |
7h00 | 17h00 | 1h00 | 5h00 | 17%/hour / 8%/hour | N/A | BA | 2D | No | Yes | here | 18 Dec. 2015 |
| JMA (Japan) |
4D-Var | 6h | 5h20 | 10h50-14h50 | -0h40 | 4h50-8h50 | 17%/hour | 4D-Var (3hr) | BA | 1D | No | Yes | here | 24 Mar. 2016 |
| Met Office (UK) |
Hybrid 4D-Var | 6h | 5h40 | 9h15 | -0h20 | 3h15 | 17%/hour | N/A | BA | 1D | No | Yes | here | 17 Dec. 2015 |
| NCEP (USA) |
Hybrid 3D-Var | 3h | 2h45 | 5h45 | -0h15 | 2h45 | 33%/hour | N/A | BA | 1D | No | Yes | here | 24 Oct. 2014 |
| NRL (USA) |
4D-Var | 6h | 6h25-6h35 | 11h10-11h20 | 0h25-0h35 | 5h10-5h20 | 17%/hour | N/A | BA | 1D | No | Yes | here | 15 Dec. 2015 |
Also, a number of centers use the Met Office assimilation system under licence.
Satellites assimilated:
| Satellite | CMA | DWD | MF | EC | ECMWF | JMA | MO | NCEP | NRL |
|---|---|---|---|---|---|---|---|---|---|
| Metop-A / GRAS | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| Metop-B / GRAS | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| COSMIC-1 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| COSMIC-2 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| COSMIC-3 | ✔ | ||||||||
| COSMIC-4 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| COSMIC-5 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| COSMIC-6 | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ |
| GRACE-A | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | |
| GRACE-B | ✔ | ✔ | ✔ | ✔(LAM only) | ✔ | ||||
| TerraSAR-X | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | ✔ | |
| TanDEM-X | ✔ | ✔ | ✔ | ✔(LAM only) | ✔ | ||||
| SAC-C | ✔ | ✔ | |||||||
| C/NOFS | ✔ | ✔ | ✔ | ✔ | ✔ |
Satellites no longer operating.
* The “loss rate” is given by: d_loss/d_latency = 100 *(1/window) where ‘window’ is the length of the assimilation window. This is an approximate measure of the percentage of data lost per hour of latency. This assumes even temporal distribution of received data, and early cut-off times that are close to the ends of the window (i.e. similar to most centres’ ‘early’ runs).
This survey is conducted by the IROWG.