Tuesday, February 26, 2013

How Important Is Lake Effect?


Our paper examining the contribution of lake-effect periods to the cool season (16 September to 15 May) precipitation of the Great Salt Lake Basin was published today in the Journal of Applied Meteorology and Climatology.  The study was funded by the National Science Foundation and the paper was lead authored by Kristen Yeager, a former University of Utah graduate student, with contributions from myself and Trevor Alcott.  It can be accessed here, but for most off campus it is paywalled, so I'll hit some of the highlights here.

Let's get right to the point.  How important is lake-effect?  We looked at some 700,000 radar images from the 1998-2009 cool seasons and identified 128 lake-effect periods (LEPs).  We then used a combination of radar and precipitation gauge data to determine how much precipitation (snow-water equivalent) fell during these periods.  The big winners were SNOTEL stations (Snowpack Telemetry stations that report daily precipitation) at Snowbird (SBDU1) in the central Wasatch Mountains and Dry Fork (DRFU1) in the northern Wasatch Mountains, both of which observe an average of about 60 mm (2.4 inches) of snow-water equivalent during lake-effect periods each cool season (left-hand panel below).  If you are wondering why Alta doesn't show up, it is simply because we didn't have access to a complete record of daily precipitation for the study period, so it wasn't considered.

Source: Yeager et al. (2013)
In the lowlands, the Salt Lake City airport (KSLC) observes an average of 16 mm (0.6 inches) of snow-water equivalent during lake-effect periods each cool season.  I suspect that if we had good data, we would find that the greatest lowland precipitation during lake-effect periods is found along the Bountiful bench and east benches of Tooele and Salt Lake County.  Unfortunately, the data at sites in those areas is skewed low because the precipitation gauges are not properly wind shielded and fail to fully catch all the precipitation during snowstorms (this is known as undercatch).  These are the sorts of warts that those of us who work with climate data have to deal with.

At the Salt Lake City Airport and Snowbird this represents 5.8% and 5.1% of the total cool-season precipitation, respectively (right-hand panel above).  Those numbers are not as large as conventional wisdom suggests, but conventional wisdom is wrong.  The lake often contributes to the tail end of a storm and, by association, it is often linked to a greater fraction of the storm total than is merited.

 The aforementioned Dry Fork SNOTEL in the Oquirrh Mountains gets 8.4% of it's cool-season precipitation during lake-effect periods, the most in the Great Salt Lake Basin.   Note that this site gets about the same amount of precipitation during lake-effect periods as Snowbird, so the greater percentage simply reflects that less precipitation is produced during non-lake-effect periods at this site.  The Oquirrhs simply don't see the same diversity of storms as the central Wasatch.

One of the more interesting findings we made was that at any given site, about a dozen lake-effect periods produce half of the total lake-effect precipitation during the 12-year study period.  In other words, at a given location, most lake-effect storms are pretty small, but a few big events make-or-break the lake-effect climatology.  Prime examples include the two lake-effect periods during the 2001 Hundred Inch Storm, which produced 107 mm of SWE at Snowbird (more than the annual average).

Finally, we examined the seasonality of lake-effect and found that it features a primary peak in the Fall (October and November), a mid-winter lull, and a secondary peak in the spring (March and April).  For skiing in the Cottonwoods, this seasonality is quite important as it means that the lake effect is concentrated in the fall when it can help build up the snowpack during the early season.  There's not much lake-effect spillover across the Wasatch Crest, so this isn't really much of a help for the Park City resorts.

Source: Yeager et al. (2013)
There is an important caveat to this work, which is that lake-effect periods sometimes feature non-lake-effect precipitation occurring in concert with lake-effect precipitation.  We didn't attempt to account for this, which is an exceptionally difficult task, and it partially explains why sites somewhat removed from the lake observe precipitation during lake-effect periods.

6 comments:

  1. Yes, a very important caveat. I can't think how one would begin to distinguish between orographically enhanced precipitation and that due purely to lake contribution. Surely there is a significant fraction where both are occurring. Do you think there is a way to quantify the error in these calculations? Perhaps this kind of study is best relagated to the digital realm. The observable world has too much uncertainty.

    ReplyDelete
    Replies
    1. Yes, this is a difficult problem. It works both ways a bit as there may be times when the lake is playing a role, but it isn't immediately apparent so the event is classified as not lake-effect.

      I think it will be interesting when we start running regional climate models with and without the lake. Of course, that will only be useful if lake effects are well simulated and the model is not strongly biased. It will take some effort to ferret that out.

      Delete
  2. It's difficult to tell where the SNOTEL sites are in the Wasatch Back. My guess is that if there were a SNOTEL site near the top of "The-Ninety-Nine-Nienty" lift at Canyons, one would detect some lake effect snow there. I have seen many powder days when it is sunny in the basin around Park City and it is still snowing while doing laps on Ninety-Nine-Ninety.

    ReplyDelete
    Replies
    1. Yes, it's not a total shutout, but in most events the spillover is limited compared to what falls on the windward side.

      Delete
  3. Do the modeling studies give you an idea of what percentage of the lake effect precipitation is from moisture evaporated off the lake, compared to how much of it is in the atmosphere to begin with?

    ReplyDelete
    Replies
    1. We did this for one event back in the day: http://dx.doi.org/10.1175/1520-0493(2001)129%3C1318:DASSOT%3E2.0.CO;2. I think, however, that we only looked at how much precip fell with and without latent heat fluxes from the lake. That's actually a bit different than tracing the water mass itself. There's always more work to do...

      My guess is that the percentage evaporated off the lake is not really large, but it could vary a lot from case to case.

      Delete