# How do they determine the weather for a specific location? **URL:** \ **Category:** Factual Questions\ **Created:** [December 15, 2017, 2:57am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554 "2017-12-15T02:57:21Z")\ **Posts on this page:** 20\ **Page:** 2
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [December 17, 2017, 5:50pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/21 "2017-12-17T17:50:28Z")
> [@Lakai](#): > > I want to know where the temperature I get when I open my app comes from. Where is the data being reported from? I get that it might be from one of the airports in the city or from central park, but how do I know which reporting site was used specifically? There is either a display or setting in the app which tells you that or there is no way to _know_. There are lots of ways to guess. But the app will decide what to display using whatever logic it wants from whatever source(s) it wants. It will either share that info with you in the UI or it won’t. And if it won’t, we’re all guessing. ---
**Author:** ![muttrox](https://avatars.discourse-cdn.com/v4/letter/m/a8b319/32.png) [@muttrox](https://boards.straightdope.com/u/muttrox)\ **Post date:** [December 17, 2017, 5:57pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/22 "2017-12-17T17:57:09Z")
Note that most of what we know as Weather Channel was bought by IBM. They do the analytics (marketing Watson heavily en route) and re-sell to all kinds of 3rd party. The TV station is the only part that was left behind, and they have tried to sell it twice. ---
**Author:** ![yabob](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/yabob/32/2821_2.png) [@yabob](https://boards.straightdope.com/u/yabob)\ **Post date:** [December 17, 2017, 7:39pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/23 "2017-12-17T19:39:59Z")
I do like Weather Underground, which has been mentioned: [https://www.wunderground.com/](https://www.wunderground.com/) They do use input from local “volunteer” stations for forecasts as well as the NWS data (which you can turn off, and only see the local volunteer station for current conditions). Slate’s writeup on it from a few years ago: > **[You Need To Start Using America’s Best Weather Forecast](https://slate.com/technology/2012/02/whats-the-best-weather-forecast-why-you-should-use-weather-underground.html)** > > If you start at the Bay Bridge and head west along most major streets in San Francisco, you’ll eventually get to a magical land of misery known as the... Apart from the accuracy and locality of their forecasts, which seems pretty good to me, I really like their 10 day forecast display, which shows a graphical timeline of up to 10 different things with about hourly granularity. ---
**Author:** ![Voyager](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/voyager/32/133_2.png) [@Voyager](https://boards.straightdope.com/u/Voyager)\ **Post date:** [December 18, 2017, 1:33am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/24 "2017-12-18T01:33:48Z")
> [@muttrox](#): > > Note that most of what we know as Weather Channel was bought by IBM. They do the analytics (marketing Watson heavily en route) and re-sell to all kinds of 3rd party. The TV station is the only part that was left behind, and they have tried to sell it twice. NBC/Universal owns it now, which is why you see shows with Today show people, NBC correspondents on some of the the bigger stories, and crap on weekends. On the West Coast they go into “Most horrible storms ever” stuff during our prime time, which is post primetime in Atlanta. Ain’t what it used to be. ---
**Author:** ![Ruken](https://avatars.discourse-cdn.com/v4/letter/r/f475e1/32.png) [@Ruken](https://boards.straightdope.com/u/Ruken)\ **Post date:** [December 18, 2017, 1:39am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/25 "2017-12-18T01:39:37Z")
> [@jasg](#): > > I don’t know what app you use but one that I have, Weatherbug has a network of basic stations (temp, baro, wind) at elementary schools in my city. It uses the one closest to my location - about a mile. Yes Earth Networks sold weather stations to schools, so you can get pretty high fidelity in some areas. They sold WeatherBug about a year ago IIRC. ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 18, 2017, 2:21am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/26 "2017-12-18T02:21:17Z")
> [@watchwolf49](#): > > I just checked … the NWS issues “spot forecasts” using latitude and longitude to 14 decimal places … that’s about one square micrometer … they used to issue forecast to 20 decimal places which is about every 8 molecules of SiO[sub]2[/sub] … you’d think they were compensating for something … > > [noparse][National Weather Service](http://forecast.weather.gov/MapClick.php?lon=-100.41174975585938&lat=41.5548086239053#.WjPouUtrwS4%5B/noparse%5D) Your comments piqued my interest in several things, primarily the significant figure issue in your attached URL. Decimal coordinates are never taken to that extreme: 13-14 characters after the decimal point… Under any circumstance. By _my_ calculations, I arrived at an equivalent area that is orders of magnitude less than yours i.e. .0111 micron squared, further increasing the ridiculous-ness factor (by orders of magnitude). But who’s math is correct isn’t the point (mine were Tequila influenced) but rather: As others have implied, the “problem” must lie in the URL coding. Interestingly, when I click on your link, sure enough, there are the offending excess of decimal places in the URL. However, when I searched the National Weather Service site myself for that same location, here’s the URL that turned up: [noparse]“[National Weather Service](http://forecast.weather.gov/MapClick.php?lat=41.9813&lon=-100.53#.Wjb8FrpFy4c)”][http://forecast](http://forecast).[/noparse] Notice the small derivation in coordinates. Additionally, every location I searched afterwards contained the same minimal 2 to 4 (hence more believable) decimal positions. Perhaps it’s a browser issue, or a DNS lookup routing difference peculiar to your locale. Who knows? It is certainly not a conspiracy by the NWS to masquerade accuracy with outrageous numbers, in a lame attempt to appease some folks’ lack of confidence in their forecasting. After the fact: I just noticed a four mile discrepancy (yours to mine) in the location descriptor on the two web pages. ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 18, 2017, 3:37am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/27 "2017-12-18T03:37:24Z")
I’ve learned so much here. Always suspected the numerous “weather apps” simply rehashed NWS data with their own twist, but I didn’t know that some utilize independent local observation stations… Interesting. I wonder if/how this improves their forecast accuracy. It seems to me, other than stating the obvious, (It’s raining on Euclid Ave. right now) unless they possess a magic algorithm that the NWS doesn’t have, this information would be pretty topical. Another interesting thing: Some comments here reinforce the impression I’ve often had of some folks when it comes to weather forecasters, i.e. they truly resent it when a prediction falls short of perfection. It makes me wonder: In what form of reality do they base their discontent, and begs the question: name any other profession that has a better track record predicting the future? Granted, sometimes it’s a bummer when they are wrong, but over the decades, weather forecasting has dramatically improved, and continues to do so. No one else can make that claim. If so, speak up… I’m listening. ---
**Author:** ![running\_coach](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/running_coach/32/15836_2.png) [@running\_coach](https://boards.straightdope.com/u/running_coach)\ **Post date:** [December 18, 2017, 3:51am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/28 "2017-12-18T03:51:15Z")
> [@yabob](#): > > I do like Weather Underground, which has been mentioned: [https://www.wunderground.com/](https://www.wunderground.com/) > > They do use input from local “volunteer” stations for forecasts as well as the NWS data (which you can turn off, and only see the local volunteer station for current conditions). Slate’s writeup on it from a few years ago: > > [What’s the best weather forecast? Why you should use Weather Underground.](http://www.slate.com/articles/technology/technology/2012/02/what_s_the_best_weather_forecast_why_you_should_use_weather_underground_.html) > > Apart from the accuracy and locality of their forecasts, which seems pretty good to me, I really like their 10 day forecast display, which shows a graphical timeline of up to 10 different things with about hourly granularity. Unfortunately, accuracy of the volunteer stations depends on the person not being an idiot. Two days ago, we had winds to 27, gusts to 38 NW(at the airport). All the volunteer stations had winds from 7-17 and some didn’t register gusts. Directions from W to SE with one SSW. ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 18, 2017, 4:01am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/29 "2017-12-18T04:01:43Z")
Not to undermine your statement but, how is the accuracy of their instrumentation a reflection on the volunteer’s intellect? Or is it something else? FYI: Our local statistics often vary greatly from our local airport two miles away. ---
**Author:** ![running\_coach](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/running_coach/32/15836_2.png) [@running\_coach](https://boards.straightdope.com/u/running_coach)\ **Post date:** [December 18, 2017, 4:28am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/30 "2017-12-18T04:28:42Z")
> [@ZurBob](#): > > Not to undermine your statement but, how is the accuracy of their instrumentation a reflection on the volunteer’s intellect? Or is it something else? > FYI: Our local statistics often vary greatly from our local airport two miles away. Flat as a pancake here. Wind is about 85-90% of the time from the NW. There’s a large flag an the entrance to the park we live in and it was clearly in a high wind on the day in question. It’s obvious that the wind gauges are not sited correctly to get a clear reading. There’s also no reason for stations a few blocks apart to have a 10 degree difference(usually summer) in temps. We don’t do microclimates here. ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 18, 2017, 5:11am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/31 "2017-12-18T05:11:37Z")
> [@running\_coach](#): > > Flat as a pancake here. Wind is about 85-90% of the time from the NW. > There’s a large flag an the entrance to the park we live in and it was clearly in a high wind on the day in question. > > It’s obvious that the wind gauges are not sited correctly to get a clear reading. > There’s also no reason for stations a few blocks apart to have a 10 degree difference(usually summer) in temps. We don’t do microclimates here. Yes, it sounds unrealistic in your situation, (no microclimates) but again… Why must it be the volunteer’s intellect at fault, as opposed to the corporation that solicited that particular inappropriate location in the first place? ---
**Author:** ![running\_coach](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/running_coach/32/15836_2.png) [@running\_coach](https://boards.straightdope.com/u/running_coach)\ **Post date:** [December 18, 2017, 5:23am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/32 "2017-12-18T05:23:00Z")
> [@ZurBob](#): > > Yes, it sounds unrealistic in your situation, (no microclimates) but again… Why must it be the volunteer’s intellect at fault, as opposed to the corporation that solicited that particular inappropriate location in the first place? From what I understand, the volunteers buy their own station and install it themselves. Wunderground does provide instructions for mounting the sensors to get the most accurate readings. [Weather station buying guide](https://www.wunderground.com/weatherstation/buyingguide.asp) [Installation guide](https://www.wunderground.com/weatherstation/installationguide.asp) ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 18, 2017, 5:52am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/33 "2017-12-18T05:52:03Z")
That changes everything. So, they are not volunteers, but rather people desiring some form of public notoriety for a fee? $$=LIKE ME.:smack: For crying out loud, no wonder these “apps” are discounted as garbage. ---
**Author:** ![LSLGuy](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/lslguy/32/5813_2.png) [@LSLGuy](https://boards.straightdope.com/u/LSLGuy)\ **Post date:** [December 18, 2017, 11:28am UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/34 "2017-12-18T11:28:48Z")
> [@ZurBob](#): > > Your comments piqued my interest in several things, primarily the significant figure issue in your attached URL. … > After the fact: I just noticed a four mile discrepancy (yours to mine) in the location descriptor on the two web pages. I’ve enjoyed the silly latlongs for quite awhile. On the NWS home page [http://www.weather.gov/](http://www.weather.gov/) if you enter your town name and state into the search box at upper left you’ll go to [http://forecast.weather.gov/MapClick.php](http://forecast.weather.gov/MapClick.php) with lat & long parameters showing silly numbers of decimals. Likewise if you go to that page and enter your zip code. The auto-suggest will come up with something like “12345, Yourtown XY, USA”. Select that and you’ll go to the same [http://forecast.weather.gov/MapClick.php](http://forecast.weather.gov/MapClick.php) with the same silly numbers of decimals. But not necessarily the same exact location since towns and zips can have different centroids. But … once you’re already on the MapClick.php page, entering either a town name or a zip code will refresh that same page with just 4 decimals in the latlongs. It looks like the same search box as the home page but whatever’s going on backstage is not quite the same. Even 4 decimal places is a bit much. That’s ~36 foot resolution in latitude and ~25’ resolution in longitude at US latitudes. Really? Such fun. ---
**Author:** ![watchwolf49](https://avatars.discourse-cdn.com/v4/letter/w/e9c0ed/32.png) [@watchwolf49](https://boards.straightdope.com/u/watchwolf49)\ **Post date:** [December 18, 2017, 5:15pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/35 "2017-12-18T17:15:20Z")
As I mentioned, this latlong was 20 decimal places a few years ago … someone at the NWS thought that was excessive and reduced to a more reasonable 14 decimal places … great work, folks, no one would think you’re compensating for anything now … > [@ZurBob](#): > > [snip] … Another interesting thing: Some comments here reinforce the impression I’ve often had of some folks when it comes to weather forecasters, i.e. they truly resent it when a prediction falls short of perfection. It makes me wonder: In what form of reality do they base their discontent, and begs the question: name any other profession that has a better track record predicting the future? Granted, sometimes it’s a bummer when they are wrong, but over the decades, weather forecasting has dramatically improved, and continues to do so. No one else can make that claim. If so, speak up… I’m listening. Weather forecasts have improved over the years, but I wouldn’t call this improvement dramatic … we’ve always had a good handle on tomorrow’s weather, and the day after … but then as today, the three-day forecast and beyond involves too much guesswork and the accuracy suffers … The problem is the lack of good data aloft, sometimes what’s happening at the surface doesn’t reflect what’s happening at 18,000 feet … it’s good we have thermometers every few miles on the ground, but here’s a map of the [NWS Rawinsonde Network](https://www.weather.gov/upperair/nws_upper) … it’s a couple hundred miles between sites that measure temperature through the entire air column … this is like asking a chemist to test a chemical with an acid without telling the chemist which acid he’s using, or how strong that acid is … Another major problem is that unsaturated air behaves differently than air that is fully saturated with water vapor … if the air is at 99.5% relative humidity we’ll see sunny weather, however if the air is 100% RH then the weather will be cloudy and rainy … just a tiny difference changes the weather completely … I think we can see the problems the average weather forecaster has relying on a humidity reading aloft from 100 miles away … this is like giving the chemist a bottle of HCl that’s been unstoppered the past three months, it’s not 12 molar anymore … The good news is that computers are getting faster by leaps and bounds … so computer modeling of weather is getting better by leaps and bounds … plus this new generation of weather satellites are just amazing in their detail … all-in-all the future is bright for forecasting accuracy, but we’re not quite there yet … ---
**Author:** ![Spiderman](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/spiderman/32/230_2.png) [@Spiderman](https://boards.straightdope.com/u/Spiderman)\ **Post date:** [December 18, 2017, 5:27pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/36 "2017-12-18T17:27:39Z")
> [@running\_coach](#): > > [Installation guide](https://www.wunderground.com/weatherstation/installationguide.asp) Most people in a city or inner ring suburb don’t have the space to meet those specs. If you get to watch a windy football game watch when they show the flags around the stadium blowing in multiple directions; the general winds aren’t doing that, the structure is what’s causing the winds to swirl. Buildings, trees, ground elevations (hills, mountains, cliffs) & even wind farms\* can cause micro wind changes. - I know of one wind farm that looks like convective activity (thunderstorms) on radar; given the forecast & other (lack of) storms showing on radar, we ignore that ‘storm cell’ when we otherwise wouldn’t fly. ---
**Author:** ![running\_coach](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/running_coach/32/15836_2.png) [@running\_coach](https://boards.straightdope.com/u/running_coach)\ **Post date:** [December 18, 2017, 7:13pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/37 "2017-12-18T19:13:33Z")
> [@Spiderman](#): > > Most people in a city or inner ring suburb don’t have the space to meet those specs. In that case, don’t hook into a network that people refer to for information. 😉 ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 22, 2017, 8:47pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/38 "2017-12-22T20:47:43Z")
> [@](#): > > =watchwolf49;20673001]As I mentioned, this latlong was 20 decimal places a few years ago … someone at the NWS thought that was excessive and reduced to a more reasonable 14 decimal places … great work, folks, no one would think you’re compensating for anything now … It appears you an I have a difference in opinion over the NWS’s competence compared to other forecasting entities. In support of your issue related to this, is the observation of their (obvious) (sometimes) decimal excess in their Lat/Long URL references. To this, I’ve suggested it is perhaps an issue of software import… To which you reject in favor as proof of a NWS _ **conspiracy…** _ I contacted (on 12/18) the NWS technical support site for an answer to this question, but as of yet, have received no answer on this issue. I’ll get back to you ASAP when they reply. > [@](#): > > Weather forecasts have improved over the years, but I wouldn’t call this improvement dramatic … we’ve always had a good handle on tomorrow’s weather, and the day after … but then as today, the three-day forecast and beyond involves too much guesswork and the accuracy suffers … We’ll just to have to agree to disagree on your statements here. Perhaps it is a matter of perspective (age). I remember when tomorrow’s forecast was sketchy at best, and day two was 50/50. Day three was in fact, a total wild guess… Now, the three day forecast is nearly a lock cinch. I consider this improvement to be very substantial. ---
**Author:** ![silenus](https://sea3.discourse-cdn.com/straightdope/user_avatar/boards.straightdope.com/silenus/32/166_2.png) [@silenus](https://boards.straightdope.com/u/silenus)\ **Post date:** [December 22, 2017, 9:39pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/39 "2017-12-22T21:39:42Z")
> [@yabob](#): > > I do like Weather Underground, which has been mentioned: [https://www.wunderground.com/](https://www.wunderground.com/) > > They do use input from local “volunteer” stations for forecasts as well as the NWS data (which you can turn off, and only see the local volunteer station for current conditions). Slate’s writeup on it from a few years ago: > > [What’s the best weather forecast? Why you should use Weather Underground.](http://www.slate.com/articles/technology/technology/2012/02/what_s_the_best_weather_forecast_why_you_should_use_weather_underground_.html) > > Apart from the accuracy and locality of their forecasts, which seems pretty good to me, I really like their 10 day forecast display, which shows a graphical timeline of up to 10 different things with about hourly granularity. Weather Underground is great. I count 14 stations in my town, and since it’s on the side of a mountain, knowing what is going on at your altitude can be important, because a mile away will be quite different. I usually center my data on the station down the block, next to the sub shop. ---
**Author:** ![ZurBob](https://avatars.discourse-cdn.com/v4/letter/z/71e660/32.png) [@ZurBob](https://boards.straightdope.com/u/ZurBob)\ **Post date:** [December 22, 2017, 10:12pm UTC](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554/40 "2017-12-22T22:12:15Z")
That is great for being aware of what is happening a mile away from you at the _present_ time (as it would be anywhere) However, how does this help in your long range planning for general forecasts? [Previous page](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554.md?page=1) [Next page](https://boards.straightdope.com/t/how-do-they-determine-the-weather-for-a-specific-location/804554.md?page=3)