Differences between animation and detailed forecast

  • I recently encountered differences of various manifestation between the animation and the detailed forecast of a location. The differences have been observed while using the wind and temperature forecast and occurred with different models but most significantly with the NEMS-model. However, the differences do not always occur and not every location seems to be affected.

    As far as I understand, the detailed forecast should be congruent with the animation and vice versa. The differences occur on Webbrowser as well as on the App.

    I am not sure but maybe my observation has something to do with this thread:

    some screenshots listed below:

    2_1532690829303_nems_wind_portocervo.png 1_1532690829303_nems_wind_marsaille.png 0_1532690829302_ecmwf_wind_portocervo.png

    2_1532690881603_nems_temp_zurich.png 1_1532690881602_nems_temp_portocervo.png 0_1532690881602_nems_temp_bastia.png

    Thank you very in advance much for having a look at this matter.

  • Sailor Moderator

    The link here below may be similar to your question:
    On the other hand, the link you presented is a different problem (wind model and waves model not matching)

  • Yes, you are right. The link you provided suits well to my observations. I am glad that this already has already been taken note of.

  • Sailor Moderator

    Not sure that the message was well recorded !
    So I want to add my own observations that I made yesterday to point out this trouble again.
    As already noted, this issue is mainly observed when using NEMS. It’s really a shame because NEMS is the only model giving good temperature forecasts in mountain.
    In the example below using NEMS, the temperature in the picker and the one in the meteogram don’t match: there is a difference of 8ºC !! While the temperature predicted by NEMS is exactly the same as the temperature (11ºC) observed in the local weather station of Les Attelas at 2700m asl.


    Same in a higher place. There is a difference of 6ºC between the picker temperature and the one in the meteogram, when the forecasted temperature (5ºC) is the same as the temperate observed in the near weather station of Jungfraujoch at 3500m asl !


    At the same place, with ECMWF model, the picker temperature and the one in the meteogram perfectly match.
    But the 11ºC forecasted are not consistent with the 5ºC observed by the local weather station. However we know that global models are not performing well in high altitude areas. .


    I hope that this will be improved soon.
    This issue is mainly noted with NEMS, and sometime with GFS and ICON.
    (I’m using Mac and iOS devices)

  • Moderator

    NEMS with problems
    0_1533301054413_2018-08-03 14_48_45-Windy_ Rüdersdorf bei Berlin weather forecast.png

    ECMWF seems to be fine
    0_1533301099042_2018-08-03 14_47_34-Windy_ Rüdersdorf bei Berlin weather forecast.png

  • Sailor Moderator

    Yes, same problem !
    When there is a big difference with the reported temperature, that takes away all confidence in the model and more in Windy. Almost nobody makes a double checking in the picker and in the meteogram.

  • @idefix37 you are right, it somehow lowers the confidence in the model and in the application itself. I don't know which "number" I should believe; the one on the map or the one in the detailed forecast.. I hope this will get a fix soon.

  • | Premium

    I just checked this issue and the difference is because on the map we display NEMS 4km, however in the detail we show Meteoblues multimodel. They are different weather models. Also, the multimodel is available globally and NEMS4km is only available in Europe.

    I will report this as a bug.

    @Tomber42 @idefix37 @tom-b @Marienka

  • @tomslavkovsky Thank you, much appreciated! Keep up the great work!

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