This website uses cookies to improve your experience. We'll assume you're ok with this, but you can opt-out if you wish. Read More
Measured performance. Clear methodology. Transparent evidence.
AlongWave regularly evaluates the accuracy of its marine-weather intelligence products against independent observations and established baselines.
This page presents the key results from the latest internal validation of our Significant Wave Height forecasts.
- The validation covered forecasts from the 1st to the 15th of February 2026, assessed over a 15-day forecast horizon using independent in-situ measurements from ten selected buoy stations in the Western North Atlantic.
AlongWave recorded lower RMSE from Lead Day 7 through Lead Day 15 in this validation.
ECMWF recorded the lowest RMSE through approximately Lead Day 6, with AlongWave showing comparable skill over the same period. From Lead Day 7 onward, AlongWave recorded lower RMSE than both ECMWF IFS and NCEP GFS across the tested Significant Wave Height forecasts.
All three forecast streams were evaluated against the same quality-controlled in-situ observations using a common harmonisation, collocation and scoring procedure.
Reference network
10 selected buoy stations from the National Data Buoy Center - NOAA
Forecast period
1–15 February 2026
Benchmark systems
ECMWF IFS and NCEP GFS
Forecast horizon
Up to 15 days/360 hours
Study area
Western North Atlantic
Same reference.
Consistent methodology.
Transparent metrics.
Forecasts were harmonised into a common validation structure and evaluated using consistent spatial, temporal and statistical criteria.
Proprietary AI-based forecasting engine
IFS
GFS
Stronger extended-range performance for AlongWave
Quantified evidence
What this validation showed
- • ECMWF remained strongest through approximately Day 6, with AlongWave showing comparable skill over the same period.
- • AlongWave recorded lower RMSE from approximately Day 7 onward.
- • The comparative advantage was strongest during Days 11–15.
- • AlongWave recorded the narrowest residual-error spread in the tested sample.
- • Extended-range error growth was more controlled.
What the validation results mean operationally
Forecast Lead time interpretation
- R² becomes low for all three systems after approximately Day 10. Extended-range guidance should therefore be interpreted as strategic planning support rather than deterministic event-level prediction.
Days 11-15 average
1,04 m
Average AlongWave RMSE
Days 11-15
Relative improvement
27,5%
lower RMSE than ECMWF IFS
19,4%
lower RMSE than NCEP GFS
Residual-error spread
Residual-error standard deviation; lower is narrower
Forecast Lead Days 1-6
Forecast Lead Days 7-10
Forecast Lead Days 11-15
Short range
Transition
Extended range
ECMWF lowest RMSE
AlongWave lower RMSE
Strongest predictive advantage
AlongWave remains competitive
Relative advantage begins
Better planning capability
Ongoing validation/calibration loop
AlongWave
0,851 m
ECMWF IFS
0,914 m
NCEP GFS
0,961 m
- For AlongWave, ongoing, transparent validation and calibration across different periods, regions and operational settings remain an active priority.