Using this app for forecasting in remote areas for skiing and climbing, one of the key pieces of information in trip planning is the freezing level (m). Is it possible to add this to line graphs for the different models? Gives us a sense of where the rain/snow lines will be and can chase the conditions.
Loving this app!
@LeCurtois Hello and Welcome to the Forum and Thank you for Using Flowx. Happy New Year
Given that Flowx depends on weather data forecasts which are notoriously unreliable, a line would need to be so wide as to be somewhat nebulous.
This comment is wrong and unhelpful Any new user searching for weather apps may look past Flowx if they read this quote.
A more correct comment would be that Flowx depends on the best weather models.
The HRRR weather model is among the best in the world with it’s 3km resolution, 15 minute time steps and a new forecast every hour!!
The equivalent in Europe is the Arome model with 1.4km resolution, 1 hour time steps but with less regular new forecasts. There is a Arome Pi model that is impressive but I haven’t added this yet.
Note, the weather models and forecasts we see is the cutting-edge of weather modelling. We spend many many billions to measure the current state of the weather and develop computer models. This is the absolute best humans have ever built.
All that is true, but even HRRR at best is a maybe.
In the past week I got soaked twice.
Last night it was raining with a predicted hour+ heavy rain. By the time I got suited up and outdoors, it was all over.
Flowx is superb app and I recommend it all the time,
Are you in a micro-climate that is not captured by the HRRR model?
Maybe you live in a micro-climate and HRRR gets it absolutely correct everywhere else??
These are more rhetorical questions to make people to think about weather models. Weather forecasts are not broad absolutes or specific incidences. Our biases notice things that go wrong rather than all the other times it’s generally correct.
I for example, find GFS reliable for planning when to mow the lawns or do gardening. I also find it very useful for watching the unstable conversion zone where hurricanes seed. Every forecast is different (i.e., wrong), but it does show the fact that the region is unstable and there is a chance a hurricane will form.
I have a motorhome.
I drive all over the western US.
NOAA NAM HRRR is ±1hr in terms of accuracy as far as rain regardless of where I am. I do as well by looking out the windows all around the house and making a guess based on the wind and cloud.
Currently:
NOAA NAM 12KM says it’s not raining
NOAA NAM 3KM says it’s raining all day. I’ve walked my dog twice in the dry. The missus got soaked this afternoon as a shower slid in from left field.
NOAA NAM HRRR 3KM says the rain has passed and it’s raining right now. Wind is supposed to be 15 knots gusting to 30. Rain is falling straight down and the flag is drooping on the pole.
The local weather from NOAA Graphical Forecast - Portland says it’s blowing 12 knots and 43% chance of rain. It’s so calm my neighbors anemometer isn’t even moving.
Flowx is my most used app and I love it!!!
You read weather very differently to me. I generally look at forecasts up to a few days out and sometimes a week or so out. You seem to be looking at current conditions.
Weather model data is more suited for forecasts rather than current conditions. In fact, weather model forecasts are used along with statistical methods (MOS) to general better forecasts for specific areas. This is what you get from most weather apps and I suspect is what you’re looking at with “NOAA Graphical Forecast”. MOS is good for local corrections but it’s limited to the local weather station(s) they use to generate the MOS forecast. So if the weather station is at the Portland airport - airport data is often used - which is usually an open area that may have 12 knot winds, and your neighbour’s anemometer is a different terrain, then the MOS forecast won’t match.
It is important to know the differences between different types of weather forecasts and their limitations.
Many people have wrong or higher expectations to what weather models/forecasts can deliver.
I do both soon and later as well.
Around home, our dog gets four walks a day and I like to have an idea of what to expect over the course of an hour or so, so I can plan when to go.
This morning it started raining about 0600 and didn’t stop 'til afternoon. NOAA HRRR said no rain until 1100.
I also like to have an idea of the expected wind as it will affect what I wear and where we go. Currently NOAA HRR says 15kn and 33kn gusts. Rain is near vertical.
NOAA NAM 3km is generally close while the HRRR is so far off it’s laughable. It purports to show a precision that it never attains. At best it’s in its Alpha stage and certainly not worth paying for.
If computer programmers were as accurate as weather forecasters, we couldn’t be having this dialog.
All the best for 2025!
Since porting to Apple which has taken many years, I haven’t kept up with the models, i.e., version and year they were updated. Maybe, the NAM model is newer than HRRR. I tried to search but found nothing.
There are differences between the models (see links below) but I wouldn’t expect major differences in accuracy otherwise they would drop the poorer model. I know NOAA develops models and compares the results to collected data. Any changes they make is analyzed for improvements before they are added to the production model. So it’s surprising to me that you are finding major differences.
https://www.reddit.com/r/meteorology/comments/o5xu4s/how_are_the_nam_conus_nest_3km_and_the_hrrr/
https://www.reddit.com/r/weather/comments/o5v9zj/differences_between_high_resolution_weather_models/
Once I finish the port to Apple and Android/Apple share the same code, I plan to work on new features and add more data sets. For example, I would love to add ensemble models - I would not be surprised if these provide some insight into our discussion.
Wow lots of banter here, and a bit confused about how any of this is related to Freezing levels haha but I’ll try to get us back on track.
In mountain weather forecasting, and in particular avalanche industry forecasting: the freezing level, Precip and winds (ideally winds at various elevations 1500m, 3000m, atmospheric) are the key pieces of information. Apart from cloud levels/Solar input. It’s all about wind, Precip, freezing level.
For anyone using this app for mountain forecasting, or to push this app into something that could totally replace SpotWx and Meteoblue, the ability to toggle winds at various elevations, as well as freezing levels would be a total game changer. Right now SpotWx seems to be used by the majority of recreational ski touring demographics who want details on their forecast.
The ski touring/climbing demographic of user groups who use these apps to plan adventures across North America and Europe is huge.
This app has an INCREDIBLE comparison GUI, the widgets are amazing, and the app is way snappier than any I have tried…but currently it lacks a few major pieces of information that lots of folks are gathering from SpotWx or Meteoblue and using it in conjunction with this app.
SpotWx does freezing levels and wind but cannot compare easily…Meteoblue does the comparisons well for Precip, but freezing levels are only on the map view and it is not so reliable for elevation winds.
These are my thoughts! Thanks for listening 
You are right, it doesn’t have much to do with freezing levels. We got on to the topic of accuracy.
All weather models are inaccurate and all weather models are accurate - it just depends on your application and definition of accurate.
I hope to do a video series on weather based on a university meteorology document I found. Hopefully, this will increase the understanding around how weather and forecasting works, and users get more value from weather models.
I do plan to add data at altitude after I finish this massive 4 year rewrite. There are some cool but complex ideas I want to try, e.g., swiping up/down levels.
Here is a blog post by Matt Ruta on ski forecasting you might like:
Thanks for the feedback. hearing about how people use Flowx guides my future designs.
Cheers, Duane.
That article Matt wrote is great, but it also fuels the point that the current forecasting options for non-professionals all only have a partial feature set for what we’re looking for so we are constantly jumping between apps.
This app is so damn close to having them all, I love it.
Thanks for all your feedback Duane!