Jump to content
  • Member Statistics

    17,584
    Total Members
    7,904
    Most Online
    LopezElliana
    Newest Member
    LopezElliana
    Joined

Big Snow threat, what will it do, part II


Typhoon Tip

Recommended Posts

Its actually the pv that is acting like a 50/50. And because its being somewhat locked in by a weak block over davis straight, its forcing the system more E then N. So really, the flow isnt as progressive as some have stated. BUT, because the pv still has to slide to through se canada still, guidance will adjust some more today into tonight. Honestly we arent THAT far off from a bigger solution.

Link to comment
Share on other sites

  • Replies 1.6k
  • Created
  • Last Reply

Could someone explain this? I hate these probability maps.

1536730_626468857400376_1886410441_n.jpg

BDL has a 90% chance of at least a 2-4 inch snowfall, an 80% chance of 4-8 inches, and 60% chance of 8-12. Usually you don't use ranges, just probability of exceeding thresholds (i.e. the probability of getting at least 4 inches, at least 8 inches, and so forth like the SREF snowfall probability maps).

Link to comment
Share on other sites

Whoever developed these probability ideas made it far more confusing than it needs to be. Just put one map out with what your forecast is

The probability based forecasts make it easier to accurately gauge forecast skill. If you just go by hard ranges, then you're either right or wrong, and your forecast skill is either 0% or 100%. If you say BDL has a 50% chance of reaching 6 or more inches, then you're going to be half right or wrong depending on whether 5.5 or 6.5 inches falls. The math in the skill scores is slightly more complicated than that, but that's the general idea behind them. Also when you consider forecast models that run let's say 50 simulations every time, you will get some distribution of forecasts (a range of predicted high temperatures, for example, with some outliers and then a tighter cluster in the middle), and probability is another way to express the composite model output.

Link to comment
Share on other sites

The probability based forecasts make it easier to accurately gauge forecast skill. If you just go by hard ranges, then you're either right or wrong, and your forecast skill is either 0% or 100%. If you say BDL has a 50% chance of reaching 6 or more inches, then you're going to be half right or wrong depending on whether 5.5 or 6.5 inches falls. The math in the skill scores is slightly more complicated than that, but that's the general idea behind them. Also when you consider forecast models that run let's say 50 simulations every time, you will get some distribution of forecasts (a range of predicted high temperatures, for example, with some outliers and then a tighter cluster in the middle), and probability is another way to express the composite model output.

It seems to me though that is the easy way out. It means you can never be wrong. To me anyway..it's kind of a way of saying..i really don't know how much we're going to get..

Link to comment
Share on other sites

Its actually the pv that is acting like a 50/50. And because its being somewhat locked in by a weak block over davis straight, its forcing the system more E then N. So really, the flow isnt as progressive as some have stated. BUT, because the pv still has to slide to through se canada still, guidance will adjust some more today into tonight. Honestly we arent THAT far off from a bigger solution.

There is also an upstream kicker that's beating down the riding out W.

Still looks good for the SE MA/ CC areas.  Nasty couple days with the cold, wind, and snow.

May not meet the "blizzard" criteria due to lower winds but visibilities will be low,

Link to comment
Share on other sites

It seems to me though that is the easy way out. It means you can never be wrong. To me anyway..it's kind of a way of saying..i really don't know how much we're going to get..

You're right in that it allows the forecaster to hedge a bit, but if you don't use probabilities scoring forecast skill becomes dependent on where you set arbitrary discrete break points on a continuous scale. For example, I could always just say the snowfall forecast for BDL is 0-12 inches and I would pretty much have a perfect forecast every time there is a storm. What if I'm the forecaster that refuses to use something other than 3 inch ranges, but there is high spatial variability such that you can't use 3-6, 6-9, 9-12, and so on without being wrong? Then what? Just keep on widening your ranges? Or try to pin down every last town's exact snowfall? That's not realistic or practical. A probabilistic forecast avoids all of that nonsense. If you want to translate a probabilistic forecast into the traditional one you're used to seeing, take the thresholds from say 75 down to 40% and that's your range. If a probabilistic forecast has 20% of reaching 12 inches, 40% of reaching 8 inches, 60% of reaching 6 inches, 80% of reaching 4 inches, and 100% of reaching 2 inches, your forecast range would be 4-8 inches as a rough way to translate between the two types of forecasts. Using a probabilistic forecast by the way implies that you're wrong at least a little most of the time, so it's hardly a way to say you're never wrong as a forecaster.

Link to comment
Share on other sites

Archived

This topic is now archived and is closed to further replies.

Guest
This topic is now closed to further replies.
  • Recently Browsing   0 members

    • No registered users viewing this page.

×
×
  • Create New...