I have to shout out for windy.com (I am not its developer sadly). It is probably the best weather site (and app) out there. You can even choose between weather models.
Advances in Weather Prediction
41–49 of 49 posts
Re: Advances in Weather Prediction
#42How good is weather prediction for other people? Is it actually accurate for you? Are even the current conditions accurate for you? I use Weather Underground, and a few weeks ago in the Bay, here are just a couple of the things I remember observing (out of the many wrong predictions): (1) One afternoon I was told there would be rain in 5-6 hours, then a couple hours I was told there would be none. (2) I started getti…
I highly recommend Darksky[1], or one of the copycat services like the AccuWeather Minutecast. They use computer vision based approaches processing radar fused with conventional forecasting, and do quite a good job. I believe that some other "hourly" forecasts are still published at the beginning of the day, rather than continually updated.
Re: Advances in Weather Prediction
#43One (rather easy?) thing I'd like to see in weather forecast apps is a confidence score. I know that sometimes the weather is hard to predict. Right now I can't tell by looking at the weather forecast how confident they are. If the different weather models give significantly different forecasts, the confidence score should reflect it.
Take a look at the probability forecast that the Norwegian met office make available as an api - https://api.met.no
Re: Advances in Weather Prediction
#44One (rather easy?) thing I'd like to see in weather forecast apps is a confidence score. I know that sometimes the weather is hard to predict. Right now I can't tell by looking at the weather forecast how confident they are. If the different weather models give significantly different forecasts, the confidence score should reflect it.
The National Weather Service office in Binghamton, NY publishes probabilistic snowfall forecasts that do a good job of breaking down the confidence scores (presumably other offices do this as well, this is just the one I'm familiar with).
There's no snowfall in the forecast at the moment, but you can see how it's broken down: https://www.weather.gov/bgm/winter
Re: Advances in Weather Prediction
#45Not widely publicized is that 5G cellular is going to eliminate one of the primary data inputs that have made weather prediction successful. The US FCC decided, on their own, that this was not an important problem, compared (most likely) to the amount of money to be made building out 5G.
>Not widely publicized is that 5G cellular is going to eliminate one of the primary data inputs that have made weather prediction successful.
When you say 5G, do you mean actual 5G in general or are you talking about mmWave (FR2, sometimes called 5G high band vs low band)? It's been frustrating having these things get mixed up, because the 5G standard has a lot of improvements aimed at more efficient spectrum utilization, further reduced latency, and other changes across a unified massive range of frequencies compared to previous standards. It should be quite useful therefore for existing spectrum as well, just as WiFi 6 brings improvements to 2.4 GHz utilization. T-Mobile for example has said its initial plans for 5G include using its 600 MHz spectrum, with mid-band and mmWave going to certain urban areas.
For whatever reason though most media and even tech people often incorrectly use 5G interchangeably with mmWave, and since 5G itself has plenty of changes it's not always clear what aspects someone is worried about. I assume in this case it's specific frequency blocks of mmWave that would be the interference concern, but I honestly don't know enough about the weather sources to be sure of that vs some other change to modulation causing more out of band interference or something like that.
Re: Advances in Weather Prediction
#46One (rather easy?) thing I'd like to see in weather forecast apps is a confidence score. I know that sometimes the weather is hard to predict. Right now I can't tell by looking at the weather forecast how confident they are. If the different weather models give significantly different forecasts, the confidence score should reflect it.
The Dutch weather forecast has an 'expert' forecast that includes confidence bands, see: https://www.knmi.nl/nederland-nu/weer/waarschuwingen-en-verw... The page is not very popular though
Thanks, never new that service existed. They should advertise it more
Re: Advances in Weather Prediction
#47Not widely publicized is that 5G cellular is going to eliminate one of the primary data inputs that have made weather prediction successful. The US FCC decided, on their own, that this was not an important problem, compared (most likely) to the amount of money to be made building out 5G.
>Not widely publicized is that 5G cellular is going to eliminate one of the primary data inputs that have made weather prediction successful. When you say 5G, do you mean actual 5G in general or are you talking about mmWave (FR2, sometimes called 5G high band vs low band)? It's been frustrating having these things get mixed up, because the 5G standard has a lot of improvements aimed at more efficient spectrum utiliza…
He's talking about 5G leaking into 23.8 GHz.
Re: Advances in Weather Prediction
#48NOAA hosts a Python SDK and REST API for historical data. And I believe they present at the SciPy conf every year (upcoming in July)
I found this talk by Uber's Danny Yuan super insightful. Forecasting is probably the subset of ML I am most excited about ;)
Two Effective Algorithms for Time Series Forecasting
Re: Advances in Weather Prediction
#49I often wondered the same thing myself, why don't weather providers provide their own accuracy rates? I started to make a "simple" system for estimating NOAA accuracy, and immediately ran into trouble determining what accuracy meant.
If they predict a high to be 86deg, and it's really 85, what does that mean as far as accuracy goes? If we use the Kelvin scale, even a 10 degree error makes it seem pretty accurate, though a person's experience in those extremes will be very different.
But I think the biggest problem is that the simple weather forecasts that we use on a daily basis, is a poor representation of what weather forecasters actually do. They're modeling how weather systems form, move, and interact. If a model predicts storm forming and moving a particular direction, but the 10 day forecast is off by 100 miles causing it to rain a day later, what does that mean for accuracy? Another model could just use the average weather as their forecast, and might score pretty high as far as long term accuracy, but would be pretty useless from a user's perspective.
So, if someone forecasts a high of 86 with a 99% confidence level. What would that mean. That it'll be 86 somewhere near there, that it'll be close to 86 at that location that day, or that it'll be 86 at that location within some timer period? You really can't boil all of those variables down into a single number.
And then you'll run into issues tracking the confidence of the confidence levels. Ad infinum.