Compare forecast models, (prediction vs performance), is pointless since past predictions are not displayed!



  • Hypothetical predictive model data utility, (accuracy), is a scientific method, (analysis), by which the boomer generation invented micro-computing, the internet, lasers, microwaves, space travel, atomic and sub-atomic particle physics, smart phones, e-commerce, digital gaming and ugh, speed- and online dating.
    OMG!
    And we always eat our brocolli.
    The angels are in the details.
    May we please have some more historical model prediction on the Windy.com weather model comparison feature?
    Thank you



  • Forecasts from old NWP model runtime would be useless to general public of weather sites. Also, keeping historical NWP grids would require huge archives, out of the scope of many web sites like Windy. Nevertheless, this remains of great interest and could be only useful to weather researchers and NWP analysts.

    By nature, forecasts are volatile: they are computed several times by day and forecasts of each new model runtime are trustworthy and more accurate than forecasts of previous model runtime (as based on more recent observations).

    When users are interested in historical (or simply previous day) data, observed values (satellite, SYNOP/METAR/TEMP/etc.) or climatological computed values are much better and useful than old forecast models.

    Anyway, I hope we'll continue eating broccoli the next 100K years :o))).


  • Sailor

    @KMWT
    Want comparison between ECMWF model (IFS) and NCEP model (GFS) performances during last years ? ...And some other weather models....
    See and click « download »
    https://www.ecmwf.int/en/elibrary/19277-evaluation-ecmwf-forecasts-including-2019-upgrade
    Good reading


  • Banned

    This post is deleted!

Log in to reply
 

Windyty, S.E. - all rights reserved. Powered by excellent NodeBB
NodeBB & contributors, OSM & contributors, HERE maps
Terms and Conditions     Privacy Policy