Wednesday, November 4, 2015

Central Wasatch Get No Love and Neither Should Icon-Based Forecasts

It's been nearly a total skunking so far for the central Wasatch in this storm.  Others have been blessed with the white stuff.  Mammoth Mountain apparently picked up 30 inches.  In northern Nevada, the Pole Creek Ranger Station SNOTEL near Jarbidge reports 4.2 inches of water equivalent in the snowpack.  Colorado is getting some today and tonight.  My twitter feed is abuzz with photos of snow from elsewhere in the west.  Mother Nature is giving us no love.  Perhaps she's upset that Ralph Becker, an avid backcountry skier, did not win the Salt Lake City mayoral election?

The hype for the next major storm has already begun as I saw it announced on the news last night.  It was done in the usual way.  The weather broadcaster says something like, "looks like a storm next Tuesday and the models are advertising a good one," with the icon for next Tuesday showing a snowflake and a maximum and minimum temperature.

Icon-based forecasts are convenient and they probably poll well in surveys of television viewers, but they are a terrible way to communicate weather forecast uncertainty and possibilities, as we've discussed previously (see Ten-Day Icon Based Forecasts Are Bogus).  In addition, although our medium-range forecast ensembles are becoming better and better at forecasting large-scale circulation changes a week in advance, snow in northern Utah is critically dependent on small-scale forecast details that are much less predictable at such lead times.

Let's take the current "storm" as an example.  We had a pretty good idea a week ago that a weather system would be moving into the western United States, with a major change in temperature.  However, there was a large range of precipitation outcomes forecast by the various ensembles, including one member that produced no precipitation.  Thus, there was a huge range of possible forecast outcomes from no precipitation to nearly 4 inches of water equivalent at Alta-Collins.  It was more likely than not that we would see a significant storm, but the odds of a skunking like we got were not zero.



So, let's once again look at the 7-day forecast for Alta-Collins.  Note how when we get out to Tuesday (10 November), we see a wide range of outcomes.  There is one ensemble member that produces nothing, most cluster between 0.75 and 1.6 inches, and one member that goes bigger.


There's little point at this time in getting all excited about this storm.  It is a synoptic possibility, not a forecast certainty.   There is a huge range of possible outcomes because of uncertainty in the location and strength of large-scale weather systems, as well as the limited predictability of smaller-scale precipitation processes that generate snowfall over northern Utah and in the Wasatch Mountains.

I'm not saying there won't be a storm in the Wasatch next Tuesday.  What I am saying is that we know right now that there is a large range of possible outcomes and my profession can do a better job forecasting and communicating the range of those outcomes and their probabilities.

15 comments:

  1. Yup, icons conceal more than they reveal. They're convenient for iphones, and so it's bandwidth that ultimately determines the quality of information accessible to most people. Weather services need to do a better job.

    ReplyDelete
  2. "What I am saying is that we know right now that there is a large range of possible outcomes and my profession can do a better job forecasting and communicating that range of those outcomes and their probabilities. "

    Summed up perfectly. Love the posts that discuss probability and predictions vs outcomes. Thank you!!

    ReplyDelete
    Replies
    1. Perhaps, but it should be "communicating the range of those outcomes..." The computer allows you to make more mistakes faster than any inventions except handguns and tequila. Will correct this.

      Delete
  3. Blogs from actual meteorologists are why I dont watch the news anymore, other then one or two guys I'm not sure any of the news guys have an actual meteorology background. What I learned this week, Anomalies are always interesting but not always fun.

    Deven

    ReplyDelete
  4. Oh dude, your such a pessimist. We're due!

    ReplyDelete
  5. I'm a broadcast met-- and I totally agree! It is frustrating to make a forecast for day 6 and 7 especially... we know about the uncertainty... but somehow we have to place some kind of number and icon in the slot... sometimes would love to go back to 5 day cast. At KSL, we try to keep it conservative with these days. Even if models scream a storm, we tend to just put in a dark cloud. Then verbally, we can explain the possibility. -G Weyman

    ReplyDelete
    Replies
    1. Grant -
      The
      private sector has helped to revolutionize weather forecasting in many ways. Although I know you are limited by the system that you have, is having to pick and present one number for a 7-day forecast really the best we can do? Tell your software providers to figure out a way to provide some level of forecast uncertainty. I have seen in other countries, for example, forecasts that include ranges, etc....

      Then again, we are one of only three countries in the world that have not adopted metric...

      Jim

      Delete
    2. Would love to see people that have meteorologist in their job profile hold an actual degree in Meteorology. How does a B.A. in music help with forecasting? You are a fine weather broadcaster.

      Delete
  6. Too bad the lake effect band hitting the northern Oquirrhs right now doesn't want to say hi to the Wasatch :(

    ReplyDelete
  7. I was in Ophir at 1:00 today. Six inches on picnic benches in town, 8 in up the canyon a bit. It was still snowing at a decent clip at 2:00 when I left. The town sits at 6500 ft, where I turned around further up canyon, maybe 6800 ft.

    ReplyDelete
  8. I could see Timp nearly all day yesterday with nothing but a dusting on the highest peaks. A disappointment for sure. One of these times we're bound to have luck with a storm.

    ReplyDelete
  9. Hi Jim, it seems one of the "critical" small scale details in breaking up this week's storm as it approached the Wasatch was the GBCZ. Can you explain how that works, or point to a link.

    Who knew?

    ReplyDelete
    Replies
    1. Unfortunately, I don't have a good blog post on it. We talk about it at http://journals.ametsoc.org/doi/abs/10.1175/2010MWR3274.1.

      The GBCZ is not always bad for precipitation in the Wasatch. It is simply a key feature involved in cyclogenesis over the Great Basin during southwesterly large-scale flow. Much depends on the details of how the system evolves.

      Jim

      Delete
  10. I think that with a high amplitude trough, the odds of this happening are a lot greater. The models tend to underestimate the more subtle types of splitting that leave us mostly dry. With the lower amplitude types of troughs I usually have much greater confidence in it hitting us.

    ReplyDelete
  11. I wouldnt say we got skunked, I think we are close to a foot for the past few days showers. Especially above 9,500 ft. Congrats on 20 years. Keep up the good work! Love your blog.

    ReplyDelete