Comments
Profiting From Weather Analytics
View Comments: Newest First | Oldest First | Threaded View
Page 1 / 5   >   >>
Re: Pick yer weather forecast model
  • 10/6/2015 3:20:15 PM
NO RATINGS

@Lyndon. Good point about the value of weather analytics. Besides the obvious life saving possibilities and having forecasts that tell us whether it's a good weekend for outdoor activities, accurate weather models can serve businesses. Agriculture of course comes to mind, but also think about retailers and how they can almost dynamically rework store displays based on forecasts, staffing plans for tourist-centric businesses, airlines that can move aircraft to areas where they won't get snowed in by a blizzard, just to name a few beneficiaries of good weather forecasting.

Re: Pick yer weather forecast model
  • 10/6/2015 3:02:58 PM
NO RATINGS

..

Jim writes


The thing I've noticed as an issue with hurricane forecasts in recent years isn't whether a storm is going to hit a particular area but when.


 

Obviously lots of opportunity for further refining of weather models, but it's amazing how accurate today's meteorological predictive analytics have become. I see this as one of the most socially beneficial areas of analytics for analytics professionals to jump into. Go Weather, Young Man, Go Weather! (and Young Woman...) (and Young Others...)

 

Re: Pick yer weather forecast model
  • 10/5/2015 12:46:29 PM
NO RATINGS

@Lyndon. The thing I've noticed as an issue with hurricane forecasts in recent years isn't whether a storm is going to hit a particular area but when. With Irene and Sandy here in the Northeast there were complaints that the models we off by something like 50 or 75 miles, so that heart of the storm didn't hit exactly where it was predicted.

My feeling is that if the model shows landfall being 50 miles away from you it's worth preparing. If it misses you in that case, so what! If it hits you, at least you were ready.

What I find frustrating is that when a storm is still in the Carribean forecasters say it is a threat to the Northeast, or to Florida, or to the Gulf, and the timeframe usually is something like "early next week". That's fine at that point. However, then the timeframe goes blank for days on end. Even the day that a storm is due to hit it seems that nobody really can say where the bad part of it is located and how fast it's moving. Should we expect it overnight? Sometime the next day? The next night? I'll never understand why that part of it is such a mystery.

 

Re: Pick yer weather forecast model
  • 10/2/2015 4:36:52 PM
NO RATINGS

..

Relevant as an update to this discussion thread...

The New York Times has an interesting article today reporting that the accuracy of the NOAA weather forecasting model has been surpassed by that of the European model:

Hurricane Joaquin Forecast: Why U.S. Weather Model Has Fallen Behind

As the article recounts,


For days, the models that guide the National Hurricane Center's forecasts had been split over the future of Hurricane Joaquin. ...

Friday, the official forecast now takes Joaquin out to sea. A direct hit on the East Coast can't yet be ruled out, but the top models doubt it.

If this forecast holds, Hurricane Joaquin will yield one clear winner: the model from the European Center for Medium-Range Weather Forecasts — or simply, the European model — which consistently forecast that Joaquin would head off to sea.


 

The article contends that the European model (as well as some elsewhere) have far greater computing capacity and thus yield greater accuracy than the USA's Global Forecast System (GFS). 

 

Re: Pick yer weather forecast model
  • 3/1/2015 10:07:12 PM
NO RATINGS

@rbaz, great point on how we use the data. It is a strong planning tool, and like all data, if used right can greatly reduce loss and improve our outcomes. 

Re: Pick yer weather forecast model
  • 3/1/2015 12:48:42 PM
NO RATINGS

@Bulk, the unexpected is what we look to have announced to us by the weather forecast. The usually normal, we anticipate and come to expect while we go about our lives. Fairly accurate forecasting enhances and enriches our activities as well as minimize inconvenience and loss.

Re: Pick yer weather forecast model
  • 2/28/2015 11:01:36 PM
NO RATINGS

@rbaz, I would say that sometimes bad weather can be just as expected. Here in Thailand between September and November, which is the rainy season we pretty much expect it to rain every single day and that there will be flooding just about every day. we just sort of plan around that. Granted, sometimes it floods for an extended period and that is never expected, but tons of rain and lots of small floods are. 

Re: Pick yer weather forecast model
  • 2/28/2015 7:21:46 PM
NO RATINGS

@Broadway, normal weather is the expected for that local. The effect any unexpected or unusual weather has on our activities is what gives the forecast such scrutiny. With all the complexities that is involved in the shaping of the weather, forecasting has its challenges. Disaster result when we misforcast the extreme and are left I'll prepared.

Re: Pick yer weather forecast model
  • 2/28/2015 10:48:21 AM
NO RATINGS

>their boats are frozen in place 

After hurricanes you often see boats sitting on top of houses and what not ... it seems like a like a lot of drama involved with boat ownership. :-)

Re: Pick yer weather forecast model
  • 2/27/2015 10:56:40 PM
NO RATINGS

That's the one thing about weather. Very few places have perfect weather ... except I guess Hawaii! But in most normal places, something goes wrong. In Texas, it's the horrible T-storms. In Boston, it's the snow. Huntsville, the tornados. Anywhere in the Gulf, horrible hurricanes.

Page 1 / 5   >   >>


INFORMATION RESOURCES
ANALYTICS IN ACTION
CARTERTOONS
VIEW ALL +
QUICK POLL
VIEW ALL +