Sunday, February 11, 2024

Snowfall Extremes at Alta, Part II

This is the second part of a blogstorm examining a new paper by Michael Wasserstein and I examining snowfall extremes at Alta.  The paper was just published in what is known as early online release and is available at, but may be paywalled if you don't work at an institution with a license for American Meteorological Society journals or have an American Meteorological Society membership (apologies).  However, we are summarizing the key findings in this blogstorm. 

In Part I, we examined the relationships between flow direction and heavy snowfall at Alta, highlighting that heavy snowfall can occur from a wide range of flow directions, but most commonly for SSW or WNW flow for Liquid Precipitation Equivalent (LPE) and the latter for snowfall amount.  The bias toward WNW flow for snowfall amount reflects that such flows are typically colder and feature higher snow-to-liquid ratios, which means you get more snow out of a unit of water.  We also identified seven key synoptic patterns for generating heavy snowfall.  

Here we examine the relationship between Integrated Vapor Transport (IVT, a common metric used to identify atmospheric rivers) and heavy snowfall at Alta.  Unlike what one might find in the Sierra Nevada or Cascades, this relationship is actually quite complex or, as I like to call it, fickle.  

To illustrate this, we looked at the time-integrated IVT (or TIVT) during extreme precipitation periods.  IVT is an instantaneous measure of the horizontal transport of water vapor over a given location.  TIVT is the total horizontal transport of water vapor over a given period (in this case 12 hours).  

The figures below present the TIVT for the snow amount and LPE extremes, with direction consistent with the 700-mb (crest level) wind and vapor transport direction.  An IVT of 250 kg/m/s, generally used as the low-end threshold for an atmospheric river, yields a TIVT of 1.1x10e7 kg/m (Editors note: This value was updated from the original post to correct an incorrect value), which is near the outer circle of these figures. These figures show that true AR conditions for a 12-hour period at Alta are extremely rare.  In addition, many snow amount and LPE extremes occur for TIVTs well below those associated with ARs.   

Source: Wasserstein and Steenburgh (2024)

So we decided to take this a step further and look at how much of the vapor that is transported over Alta is converted into precipitation during the snowfall amount and LPE extremes.  We call this the Local Precipitation Efficiency.  Consistent with the analysis above, this efficiency is especially high during WNW or NW flow. 

Source: Wasserstein and Steenburgh (2024)

So those northwesterly flow storms get a lot out of a little.  

Finally, we decided to look at what happens when the IVT is high at Alta.  Given that true AR conditions with IVT ≥ 250 kg/m/s are extremely rare at Alta, we lowered the IVT threshold to 200 kg/m/s.  During the 23 cool seasons we examined, there were 112 periods (about 5 per year) in which the mid-point IVT was ≥ 200 kg/m/s.  Of these, only 19 produced an LPE extreme at Alta and 37 produced no precipitation at all.  These are the the highest local IVT events a year although some can go big, some are total busts.  

All of this indicates that one needs to avoid what I'll call AR-myopia at Alta.  Snowfall extremes, especially in NW flow, can occur with relatively low IVT.  Locally high IVT (≥ 200 kg/m/s) can sometimes produce a big snowfall amount or LPE event, but there are times when it produces little to no precipitation.  The correlation between IVT and precipitation at Alta simply is not high enough to justify using IVT in isolation for precipitation prediction.  

To that point, we also breakdown the differences between high IVT events that produce an LPE extreme and those that don't produce LPE.  The former typically are colder, feature higher relative humidities, and stronger large-scale ascent.  The latter is sometimes referred to by meteorologists as "forcing" and would be produced by, for example, and upper-level trough and/or surface front.  Essentially, you need an environment that favors precipitation generation.  During high IVT, forced ascent over the central Wasatch, by itself, is not sufficient to do the job if the airmass is not close to saturation and there is a lack of large scale forcing for precipitation.   

Thanks for reading!


  1. Great work Michael — I’m curious if you could speculate on the following. If large scale forcing is required for extreme precip, I wonder what about the terrain at Alta favors such extreme snowfall accumulation climatologically? I’ve always assumed it was because the terrain is oriented favorably w.r.t to the typical winter IVT flux. But this model suggests that the large scale forcing is more (?) important, so I’m wondering how terrain interactions fit into that picture if not by mechanical uplift. Maybe seeder feeder processes are well suited to happen here versus other nearby ranges? Sorry for the rambling question

    1. Jim here (not Michael). The synoptic environment in high IVT events is very different from post-frontal NW flow situations. The high IVT events are often more stable and less frequently feature convection. In contrast, the NW flow events feature what meteorologists call potential instability, which results in the initiation of convection when the flow is forced over the mountains.

      Additionally, IVT can come in many distributions over the western interior. I did not discuss this in the post, which already too long, but in the paper we highlight differences between the environments of high IVT events that produce extreme precipitation and those that produce no precipitation.