Dealing with inaccuracies
-
Hi,
I Love the Windy products but occasionally there's some significant errors in the current winds and waves. Also, would suspect errors in the forecasts to but that's hard to predict. Would like to get info to provide better info, perhaps a different model or something.I currently use this for boating in the Eastern US, coastal waters, great lakes, and gulf of Mexico. I'm also an aviator so most of the weather I use is aviation weather.
If one compares the Windy Winds and Waves with a current known weather station, there are differences of up to 40 to 50% in both waves and winds.
The current stations include observations from bouys and airports, which are very accurate and timely... from a few minutes to an hour at the most.
Im using the Wavewatch model for waves. Where does this data come from? Assume it uses bouys? The only other option is ECMWF which doesnt seem as accurate. How do you get forecast wave height and period?
For the winds, I use NAM, but seems like GFS is pretty close. The NAM doesn't specify CONUS, but would assume that what it is because I'm in the US, correct?
I find it quite useful for general planning, but would like to have the right settings to get a bit better info.
Thx much for your help.
Seevee
St. Pete, FL -
@seevee
You talk about errors in the current wind or waves and in their forecast on Windy. If you seek for better NWP models than those used by Windy, it’s not here that you will find an answer.
ECMWF model and GFS are considered to be the best models specially for 5 - 10 days forecasts.
Concerning Wavewatch model, it has been developed by NOAA /NCEP:
http://polar.ncep.noaa.gov/waves/wavewatch/
ECMWF use WAM which show very low error for wave height - and peak period too -(The error between forecasted and observed values is expressed by the scattered index)
https://en.m.wikipedia.org/wiki/Wind_wave_model
NAM has a better resolution and can be useful for short term prediction for wind and weather.
Then there are plenty of information about Numerical Weather Prediction. You may start with https://en.m.wikipedia.org/wiki/Numerical_weather_prediction
-