Wave model verification:
Predicted significant wave height vs. buoy data / satellite altimetry
This page is updated every 6 months (january/july)
Dansk version

Model, data, and statistical methods
Satellite data


Model vs. buoys - on-line validation

A simple, non-quality checked, data display

Most recent data

Ensemble forecast

Previous month
Two months back

Wave spectra

A simple, non-quality checked, forecast data display

1-d spectra
2-d spectra

Model vs. buoys - 6-monthly validation

Detailed comparison of 6 months of buoy data and operational forecasts of significant wave height. The general wave height forecast quality is evaluated for each buoy. The wave height during the 5 most severe storm events is studied separately. Stations are intercompared, and the dependency of the wave height forecast quality on both forecast range and observed wave height is calculated. The quality Q is based on the 5 highest events at a fixed, small subset of wave buoys.

Previous half-year
Previous 4 years, compared
Highest waves

Jcomm: Model vs. buoys - monthly

Jcomm (Joint WMO-IOC Technical Commission for Oceanography and Marine Meteorology) carries out an automatised, global comparative wave model verification. DMI takes part in this project, and submits wave forecasts for a number of buoys to ECMWF once every month. Please refer to

Jcomm verifikation - rapport
Jcomm verifikation - grafik

for a project description. Scroll to "Automatic Documents" for monthly reports.

Model vs. satellite altimetry

Comparison of 6 months of satellite altimetry based and operational forecasts of significant wave height. The forecast error is mapped for the Northwest European Shelf Seas, and mean errors are calculated as a function of wave height, forecast range, and time of year.

Last year
Previous 3 years, compared

Previous studies

Validation reports since 2001


Revised depth maps

The impact of updated depth maps, with ETOPO1 / ETOPOP2 replacing ETOPO5 as data source, is studied in a hindcast simulation of 2010, jan-jun.

Previous half-year
Highest waves

Current refraction

The refractive effect of time-dependent current, as calculated by the DMI operational ocean hydrodynamical ocean models, is included.

Previous half-year
Highest waves
Test/opr model intercomparison

Extended spectral frequency range

Attempt to entend the high-frequency range of the modelled spectrum.

Test/opr intercomparison

Case Studies

Comparison with buoy data

Wave forecasts at wind mill parks, primo 2009
Comparison of 3 different WAM version in the shallows of wind mill park sites for a 1½ month period.

Wave forecasts off Ireland, ult. 2003
Preliminary study of forecasts for 3 Irish buoys during a 3-week period.

North Baltic Proper, jan. 2005
Hurricane over the Baltic, with exceptionally high waves.

Comparison with satellite altimetry data

North Sea - Baltic, med. 2004 - mar. 2005
Sea state as measured from satellite is described and compared with wave model forecast. The model error (bias, stdev, scatter index) is mapped for all European waters excluding the Mediterranean. The area-averaged model error is given as a function of time, forecast range, and wave height.

Yellow Sea , medio-ultimo 2008
Same, for Chinese and Korean waters.

Yellow Sea , medio-ultimo 2008
Same, for Chinese and Korean waters. Extended model area.

Yellow Sea , primo-medio 2009
Same, for the Red Sea.

South China Sea , primo-medio 2009
Same, for the South China Sea.

Jacob Woge Nielsen, Jacob L Høyer - May 14th 2018