One of the most important aspects of a weather system is undoubtedly the quality and accuracy of the weather forecast.
Although there are numerous weather providers active in this market there is an enormous difference in the forecast quality provided.
The main differentiation is the sophisticated methods used by Meteo Consult in order to ensure that the weather information provided is accurate in comparison with the verifications of other providers.
In 2005, the Meteo Group developed a unique forecast system for one of their major offshore clients, ‘Shell’. This system called NMB was implemented in the forecast operations of about 100 shell platforms. Other major operations like ExxonMobil also use these forecasts as a result of their proven reliability.
The best global forecasting models available today are:
- ECMWF of the joint European weather service
- UKMO or ‘Bracknell’ model of the UK Meteorological Office
- NCEP of the American National Weather Service (part of NOAA)
Other maritime weather software makes use of these models.
Each model has it’s own qualities. Meteo Consult calculates a new forecast by weighing these three models up with the Ensemble forecasts of ECMWF and then making a projection. Unique is that the reliability of the forecast is directly taken into account. We call this model our Nautical- Meteo- Base (NMB).
MeteoGroup NMB model out performs the above mentioned models (see table 1 and 2) in continuous verifications. By using 3 models as input, the reliability of delivering the forecasts on time has improved as there is no single dependency.
NMB | ECMWF | UKMO | NCEP | |
Day 1-3 | 2.47 | 2.70 | 2.75 | 2.98 |
Table 1: Mean Absolute Error (MAE) for forecasted wind speed in knots, for days 1-3 ahead, 2 month verification (September/October) for 9 locations.
Since 2006, NMB data has also been implemented in SPOS. In addition to pressure systems, wind and waves other elements like visibility, risk of icing, weather (snow, rain, fog, etc.) and ice edges, ocean currents and tropical cyclones have also been added.
There is no doubt that NMB makes the SPOS forecasts the best available on the market.
ECMWF | UKMO | NCEP | |
Day 1 | 5% | 9% | 20% |
Day 2 | 8% | 11% | 21% |
Day 3 | 13% | 14% | 21% |
Day 4 | 16% | 17% | 27% |
Day 5 | 20% | 11% | 24% |
Day 6-8 | 16% | - | 29% |
Table 2: Out performance of NMB compared to individual input models
How to read: forecast for day 4 ahead: ECMWF forecast error is 16% larger than NMB forecast error, NCEP error is 27% larger than NMB