Slashdot Mirror

The weather reports are worthless. The dudes that give them to us are basically just flimflam men. If we got our weather predictions by rolling dice, they'd be just as reliable. I say we stop paying for this nonsense and tell the meteorologists to get a real job.

I agree with your criticism of the GP's post. Weather reports are pretty bad, and they aren't really getting much better. Nevertheless, they aren't quite as bad as you imply. If you'd like to get a sense of the relative accuracy for various weather forecasting organizations in your area, you might want to check out:

http://www.forecastadvisor.com/

There have been studies on this, and depending on how you interpret the data, your local TV weatherman is probably useful for 24-36 hours of forecasting. (Some really aren't much better than chance, but most can at least give a prediction somewhat better than chance for the weather on the next day.) Beyond that, you'd be better off consulting some national service. The error rate gets bigger and bigger, but basically once you get somewhere between 5-8 days out, the predictions really aren't any better than chance. So don't believe 10-day or 14-day forecasts -- they are just nonsense. Basically, for anything more than a few days out, consult climate data for your region and go from that.

In any case, weather prediction isn't that great, but it is better than chance, at least for short time frames.

Intellovations (my company) does collect lots of forecasts including probability of precipitation forecasts. One was even published by CustomWeather.

There are two measures of POP forecasts: accuracy, and resolution. The published study only measured accuracy. Measuring accuracy means verifying it rained 20% of the time the forecast said 20% chance of rain. Resolution means how useful the POP forecasts are. Since it it rains/snows about 30% of the time, always predicting 30% chance rain or snow is 100% accurate, but useless. A perfectly useful POP forecast would always predict either 0% or 100% for any given point. A perfectly useful and accurate POP forecast would always predict either 0% or 100% for any given point, and be right 100% of the time. Real POP forecasts fall somewhere in the middle.

You can check out more at the sites ForecastAdvisor.com and ForecastWatch.com.

Intellovations (my company) provides a weather forecast accuracy calculation product that is used by many meteorologists. It also compares forecasts against both climatological (tomorrow will be average) and persistence forecasts (tomorrow will be the same as today).

I can tell you that skilled weather forecasts handily beat persistence forecasts overall. However, there are some specific circumstances where that isn't true.

One other interesting thing is that in general persistence forecasts are more accurate than climate forecasts out to about 3 days, when they become less accurate than climate. That says a little something about how long weather systems last.

Check out the sites at ForecastAdvisor.com and ForecastWatch.com for more.

I was quite curious about weather forecast accuracy as well. So three years ago I started collecting weather forecasts from the primary providers (Accuweather, Weather Channel, NWS, CustomWeather, Intellicast, etc.) and comparing them to actual observations. It's tougher than you might imagine, and there are a lot of factors that need to go into creating usable verification statistics.

I have a public site with some statistics for about 800 locations in the US available at ForecastAdvisor.com. There is also a blog with more in depth analysis (like how do temperature forecasts fare relative to how deviant the actual temperature is...in other words how well do forecasts do the further away from normal the actual is, and how to forecasts fare the further out they forecast for, and how does forecast accuracy compare over time.).

ForecastWatch.com is used by meteorologists and professionals. Accuweather, The Weather Channel, and several private meteorological companies use this system to help them understand and improve their weather forecasts.

And a geek note: ForecastWatch.com runs on Quixote (a Python web framework), while ForecastAdvisor.com runs on Ruby on Rails. The back-end forecast and actual collection, and calculations are Python with a MySQL database. Both sites are close to migrating to Django, a new Python web framework and ORM.

-Ace

I was quite curious about weather forecast accuracy as well. So three years ago I started collecting weather forecasts from the primary providers (Accuweather, Weather Channel, NWS, CustomWeather, Intellicast, etc.) and comparing them to actual observations. It's tougher than you might imagine, and there are a lot of factors that need to go into creating usable verification statistics.

I have a public site with some statistics for about 800 locations in the US available at ForecastAdvisor.com. There is also a blog with more in depth analysis (like how do temperature forecasts fare relative to how deviant the actual temperature is...in other words how well do forecasts do the further away from normal the actual is, and how to forecasts fare the further out they forecast for, and how does forecast accuracy compare over time.).

ForecastWatch.com is used by meteorologists and professionals. Accuweather, The Weather Channel, and several private meteorological companies use this system to help them understand and improve their weather forecasts.

And a geek note: ForecastWatch.com runs on Quixote (a Python web framework), while ForecastAdvisor.com runs on Ruby on Rails. The back-end forecast and actual collection, and calculations are Python with a MySQL database. Both sites are close to migrating to Django, a new Python web framework and ORM.

-Ace

I was quite curious about weather forecast accuracy as well. So three years ago I started collecting weather forecasts from the primary providers (Accuweather, Weather Channel, NWS, CustomWeather, Intellicast, etc.) and comparing them to actual observations. It's tougher than you might imagine, and there are a lot of factors that need to go into creating usable verification statistics.

I have a public site with some statistics for about 800 locations in the US available at ForecastAdvisor.com. There is also a blog with more in depth analysis (like how do temperature forecasts fare relative to how deviant the actual temperature is...in other words how well do forecasts do the further away from normal the actual is, and how to forecasts fare the further out they forecast for, and how does forecast accuracy compare over time.).

ForecastWatch.com is used by meteorologists and professionals. Accuweather, The Weather Channel, and several private meteorological companies use this system to help them understand and improve their weather forecasts.

And a geek note: ForecastWatch.com runs on Quixote (a Python web framework), while ForecastAdvisor.com runs on Ruby on Rails. The back-end forecast and actual collection, and calculations are Python with a MySQL database. Both sites are close to migrating to Django, a new Python web framework and ORM.

-Ace

I was quite curious about weather forecast accuracy as well. So three years ago I started collecting weather forecasts from the primary providers (Accuweather, Weather Channel, NWS, CustomWeather, Intellicast, etc.) and comparing them to actual observations. It's tougher than you might imagine, and there are a lot of factors that need to go into creating usable verification statistics.

I have a public site with some statistics for about 800 locations in the US available at ForecastAdvisor.com. There is also a blog with more in depth analysis (like how do temperature forecasts fare relative to how deviant the actual temperature is...in other words how well do forecasts do the further away from normal the actual is, and how to forecasts fare the further out they forecast for, and how does forecast accuracy compare over time.).

ForecastWatch.com is used by meteorologists and professionals. Accuweather, The Weather Channel, and several private meteorological companies use this system to help them understand and improve their weather forecasts.

And a geek note: ForecastWatch.com runs on Quixote (a Python web framework), while ForecastAdvisor.com runs on Ruby on Rails. The back-end forecast and actual collection, and calculations are Python with a MySQL database. Both sites are close to migrating to Django, a new Python web framework and ORM.

-Ace

I was quite curious about weather forecast accuracy as well. So three years ago I started collecting weather forecasts from the primary providers (Accuweather, Weather Channel, NWS, CustomWeather, Intellicast, etc.) and comparing them to actual observations. It's tougher than you might imagine, and there are a lot of factors that need to go into creating usable verification statistics.

I have a public site with some statistics for about 800 locations in the US available at ForecastAdvisor.com. There is also a blog with more in depth analysis (like how do temperature forecasts fare relative to how deviant the actual temperature is...in other words how well do forecasts do the further away from normal the actual is, and how to forecasts fare the further out they forecast for, and how does forecast accuracy compare over time.).

ForecastWatch.com is used by meteorologists and professionals. Accuweather, The Weather Channel, and several private meteorological companies use this system to help them understand and improve their weather forecasts.

And a geek note: ForecastWatch.com runs on Quixote (a Python web framework), while ForecastAdvisor.com runs on Ruby on Rails. The back-end forecast and actual collection, and calculations are Python with a MySQL database. Both sites are close to migrating to Django, a new Python web framework and ORM.

-Ace