NOAA Privacy Policy | NWS Disclaimer
Link to National Oceanic and Atmospheric Agency WW3 validation logo Link to National Weather Service

AKW Model Validation

 Point of contact: NCEP.EMC.wavewatch@NOAA.gov
Last update: October 15, 2004
MMAB page | Main wave model page
 
NWW3, AKW or WNA validation page
This page presents error statistics of the regional AKW wave model and the wind fields driving this model. In set-up and data used it is similar to the validation page of the global model. Details on data used and validation methods can be found on the latter page. Only differences will be discussed here some detail.

Wind Validation

Buoy Data

In the analysis of the buoy data two groups of buoys are considered. The first are the deep ocean data points that are also used in the global NWW3 model. The second group are the coastal buoy locations, that are not present in the global model. The data analysis is similar to that for the global model, resulting in:
Remarks on the above Figures:
  • Note that the wind speed errors for the coastal buoy locations are larger both in absolute and relative sense than those for the deep ocean buoys. Two possible reasons for this are the mixture of over-land and over-sea characteristics in the GDAS and AVN for these points, as well as.

Satellite Data


The validation with the satellite data is performed in the same way as for the global model, with two differences. First, the validation is performed with the altimeter data at native resolution. Secondly, it is collocated on a 0.5O resolution grid. Both differences mean that the comparison includes smaller (meso-) scales in the data. Therefore (i), rms differences and scatter indices against this data may be expected to be larger for identical model wind fields, and (ii) the validation plots will generally be less smooth.

mean observed U10 from altimeter
hindcast vs. altimeter
24h forecast vs. hindcast
48h forecast vs. hindcast
72h forecast vs. hindcast
1999/09 - 1999/11
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
1999/12 - 2000/02
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/03 - 2000/05
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/06 - 2000/08
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/09 - 2000/11
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/12 - 2001/02
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/03 - 2001/05
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/06 - 2001/08
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/09 - 2001/11
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/12 - 2002/02
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2002/03 - 2002/05
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2002/06 - 2002/08
N.A.
N.A.
N.A.
N.A.
2002/09 - 2002/11
N.A.
N.A.
N.A.
N.A.
Comments on Figures in the above Table:
  • See NWW3 validation page regarding interpretation of these figures.

Top of page | Main wave model page

Wave Model Validation.

Buoy Data

The wave data from the buoys is separated into groups and analyzed in the same way as the wind data.
Comments on above figures:
  • Before December 1999 systematic negative biases occurred due to an error in the boundary data transfer into this model
  • Results for this model for deep ocean and coastal are very similar. This is probably due to the fact that even for the coastal buoys the dominant wave fields are generated in the deep ocean and travel towards the coast

Satellite Data

The validation with the satellite data is performed in the same way as for the global model, with two differences. First, the validation is performed with the altimeter data at native resolution. Secondly, it is collocated on a 0.5o resolution grid. Both differences mean that the comparison includes smaller (meso-) scales in the data. Therefore (i), rms differences and scatter indices against this data may be expected to be larger for identical model wind fields, and (ii) the validation plots will generally be less smooth.

mean observed Hs
hindcast
24h forecast
48h forecast
72h forecast
1999/09 - 1999/11
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
1999/12 - 2000/02
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/03 - 2000/05
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/06 - 2000/08
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/09 - 2000/11
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2000/12 - 2001/02
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/03 - 2001/05
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/06 - 2001/08
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/09 - 2001/11
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2001/12 - 2002/02
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2002/03 - 2002/05
bias / rms / SI
bias / rms / SI
bias / rms / SI
bias / rms / SI
2002/06 - 2002/08
N.A.
N.A.
N.A.
N.A.
2002/09 - 2002/11
N.A.
N.A.
N.A.
N.A.
Comments figures in above Table:
  • Large negative biases close to the open boundaries for 1999/09 - 1999/11 indicate problems with boundary data transfer into the model. For subsequent periods this error disappears
  • In spite of the higher resolution of this model, biases still indicate problems with resolving the Aleutian Islands in this model
Top of page | Main wave model page
NCEP | National Weather Service | NOAA | Department of Commerce