Hacker News Reader

Comment by counters | Hacker News Reader

Do you want a snarky answer or a serious answer?

The serious answer is that the way you'd try to figure this out is by combining weather radar, satellite imagery, and a nearby surface observation to try to estimate the current conditions. But there can be a latency of up to a few minutes from these sources, and they could disagree with one another. You have to use them to bootstrap your near-term or nowcast product, but enforcing consistency with recent real-time and the nowcast is quite hard.

It's a surprisingly nuanced technical challenge. Most of the time, it works out just fine (e.g. if there is no weather). But people are awfully good at remembering when these sorts of analyses end up being wrong!

View on HN

Yeah. Like you would think you could just look at reflectivity data to determine whether or not it's currently raining, but at most places you are far from a radar site and even the 0.5 degree tilt is scanning a mile above your head. There might be rain there, but is it reaching the ground? All you can really do is guess.

If you're interested in providing on-the-ground condition reports, install mPING: https://mping.nssl.noaa.gov/

I keep this app on my homescreen and try to report when very light rain starts, since it's not always obvious from the reflectivity data. Ultimately the user reports get fed into things like improving the model, and more data is always good.

Apple devices are constantly phoning home every time they see a random AirTag out in the world.

You'd think that if their users are accepting that level of communications with the mothership that they could ship some AI model to hear rainfall in the wild, and thus improve their live weather data.

Surprisingly low signal-to-noise ratio for most of the common, creative ways people come up with to detect rain. Windshield wipers on cars are another example.

The thing is, even if you did have a super reliable in situ "rain detector", how do you combine it with the existing datasets like weather radar, which is a gridded product? This is actually a really, really difficult sensor fusion problem when you then super-impose product requirements like the general location real-time detection map and the inputs necessary for whatever internal nowcasting system they use.

If you care that much buy a $100 weather station and sit it outside your window?

Actual ground observation weather stations are fairly rare outside of places like airports and major news stations.

“Is it raining here right now?” Is a harder question than you’re giving it credit for. Radar can show rainfall at as low as a couple thousand feet altitude, but if conditions are right/wrong (depending on how you look at it) it never reaches the ground.

What's wrong with devices phoning home if they don't actually send any usable data there?

I believe they send barometer information home

The AirTag comparison is interesting. AirTag tracking is clever and results in Apple not knowing the location or identity of either the AirTag or the phone reporting it. This relies on rotating keys that can be seen as random, or at least not identifiable.

But rain or other objective information? I suppose it works, maybe a bit like “limit IP address tracking” — cloudflare or other edge provider could mediate so Apple gets the data, knows it comes from an iPhone (to prevent bad data attacks), but Apple can’t tell what phone sent which data.

(the privacy concern being documentation of when you were inside/outside/etc).

> if there is no weather

And the ship has been towed beyond the environment.

There is nothing out there, all there is is sea, and birds, and fish. And 20,000 tons of crude oil. And a fire.

Less sarcasticaly speaking I think there is always weather. Maybe what you mean is “no significant change in the weather” neither in time, nor in space.

It would be kind of interesting if the app had a “you are wrong” button, which allows you to take a picture of the outdoors. Apple could either use this to improve their models, or even just use it as input data directly if they get enough complaints. Plus, it would allow people to vent, or it could check if there is something wrong with the phone, maybe location is being mis-read or something like that.

It does have that: “report an issue.”

There isn't a vector where after-action reports like this could "improve the model." That data is useful for verification, but these systems generally have no learning component to feed the data back into them to improve them.

They have the input from all the different sensors and forecasts. Why not, for a given location, keep track of which one gave the best results? Sure, they don’t have a way of keeping track of they now, but it seems like it could be added.

This is already SOP at most reputable weather data providers; they consume many different numerical forecasts and use statistical post-processing to choose an optimal blend of the available forecasts based on how different forecasts have verified against observations.

But this sort of technique only works for medium range forecasts. Short-range precipitation nowcasts are almost always a single, deterministic run of a model that extrapolates from patterns in recent radar imagery. They aren't bias corrected at all, so you can't use observations in the same way to improve them.

Yeah - nowcasting turns out to be remarkably difficult at times, especially at very small spatial resolutions.

Wait, what's the snarky answer?!

Reply on news.ycombinator.com