Wednesday, December 7, 2016

A Primer on Atmospheric Rivers

Since the first atmospheric river event of the season will be affecting northern Utah later this week, it seems appropriate that we spend a few minutes today discussing just what the hell an atmospheric river is, how they penetrate into the western U.S. interior, and what sorts of impacts that they have on snow and skiing in the Wasatch Range.

What is an atmospheric river?

Atmospheric rivers are narrow corridors of strong vertically integrated water vapor transport.   Vertically integrated means that we are interested in the total water vapor transport through the troposphere, the lower weather producing portion of the atmosphere that extends to about 15 km above sea level (give or take).  The phrase was first coined by MIT scientists Reginald Newell and Yong Zhu in a 1994 research paper, although they used the term tropospheric river for the same phenomenon in an earlier paper.  Their major contribution was recognizing the importance of these features for understanding the global water budget, the export of moisture from the tropics to the mid latitudes, and the generation of precipitation in the mid and high latitudes (their 1994 paper was particularly interested in understanding Greenland ice-core records).

Prior to the 1990s, atmospheric rivers were known by various names.  For example, western US meteorologists used the term Pineapple Express to describe moisture plumes originating from near Hawaii.  There's even an IPA brewed by Drought Works in Missoula, MT with the name.

Most atmospheric rivers originate in from the tropics or subtropics and are often found ahead of cold or occluded fronts associated with midlatitude cyclones.  Their high water vapor content reflects their origin in the tropics, but also the convergence of mid-latitude moisture, which can offset water vapor losses to precipitation.

Ideally, atmospheric rivers are identified using a diagnostic quantity known as integrated vapor transport, or IVT, since it is the transport of water vapor that is most important and also correlates best with mountain precipitation.  In practice, integrated water vapor, or IWV, is often used since it is more readily measured by some satellites and a common field in model output grids.  We have found IVT (figure b below) to be a better diagnostic than IWV (figure a below) for tracing atmospheric rivers inland over the western US.  It also has a stronger correlation with precipitation, especially in the mountains.

Source: Rutz et al. (2014)
Strong objections to the term atmospheric river have been made by some scientists, in part because their behavior is so different from water rivers.   I have mixed views, but feel the term is here to stay and think it has drawn the attention of the public in a way that would not have been possible without such a catchy phrase (although this has downsides too).

Inland Penetration

Atmospheric rivers typically weaken as they move inland into the interior of the western United States.  Precipitation, especially over mountain barriers like the Cascades and the Sierra Nevada, removes water vapor from the river, weakening the integrated vapor transport.  As a result, inland penetration more likely if the river is strong when it makes landfall and if it traverses areas with lower topography.

The high Sierra Nevada south of Lake Tahoe is an atmospheric river graveyard as it is the highest barrier in the Pacific states.  If one looks at low-level trajectories launched within atmospheric rivers when they make landfall, those that make it into the interior within an atmospheric river nearly always circumscribe or avoid the southern high Sierra.  Atmospheric river conditions are extremely rare downstream of that high barrier, especially over central Nevada and western Utah.  Northern Utah and the Wasatch Range are fortunate to be downstream just enough to benefit from river that circumscribe the range.

Source: Rutz et al. (2015)
Thus, the preferred pathways for atmospheric river penetration into northern Utah are: (1) a westerly, west-northwesterly, or weakly clockwise curving trajectory crossing the lower terrain of the Sierra Nevada and Cascades of northern California and Oregon or (2) a southwesterly or weakly counter-clockwise turning trajectory across Southern California or the Baja Peninsula and following the lower Colorado River Valley.

Impacts on snow and skiing in Utah

On average, bonafide atmospheric river conditions exist over northern Utah on only a few days each cool season (October to March), although atmospheric river remnants certainly contribute to precipitation more frequently than that.  Not all atmospheric river events are huge snow producers.  Much depends on the flow, instability, large-scale characteristics, and duration of the event.  However, some produce prolific amounts of snow and water equivalent.  Others can feature high snow levels (bummer!).   I tend to look forward to major, long-lived atmospheric river events as they are nearly always noteworthy, often provide memorable skiing (sometimes good, sometimes challenging), and sometimes memorable avalanche cycles.  For example, in my book Secrets of the Greatest Snow on Earth, I discuss the 17–23 December, 2010 atmospheric river event that produced 75 inches of snow at Sundance Ski Area, snow levels that reached as high as 7500 feet, and massive avalanches on Mount Timpanogos.

The 17–23 December 2010 atmospheric river event.  Photos courtesy Bill Nalli.
Later this week

Forecasts from the Global Ensemble Forecast System (mean IVT top figure below) and GFS (bottom figure) suggest we will see atmospheric-river or near-atmospheric-river conditions pushing into northern Utah late Thursday and persisting through Saturday morning.  Meteorologists use a threshold (red line below) for identifying atmospheric rivers (red line below) and the forecast for 0000 UTC Saturday (5 PM MST Friday) shows how the atmospheric river is able to penetrate inland across the lower northern Sierra Nevada and Southern Cascades.  Note also how the water vapor transport declines rapidly across the southern High Sierra.

Water totals being spit out by the models and ensembles range from 1 to 3 or more inches through Saturday, depending on location.  If time permits, I hope to take a closer look at the forecasts tomorrow.  

Addendum @ 9:20 AM

I forgot to include the all important summary figure showing the preferred regimes and pathways for atmospheric river penetration into the western U.S. interior.  Much thanks to U alum Jon Rutz for leading this work.

Source: Rutz et al. (2015)


  1. The rain/snow line has got my attention big time. You indicated yesterday the LCC guidance wasn't that reliable 48/72 hours out. I've come to prefer the NAM4 for snow/precip, but don't know about its wetbulbzero estimates. For the sake of conversation it increases from 4400 ft 5am Fri to 7900 ft 11am Fri, drops to 6800 ft 2pm Fri increases to 8000 ft 5pm Fri. Corresponding rain/snow line is maybe 1000 feet lower. Seems like you take this w a grain of salt. Still, can we conclude rain/snow line will rise throughout the day Friday, perhaps reaching 7000 ft or higher?

    Is the reason for the rain/snow mix simply that the storm originates near Hawaii in the sub-tropics, hence Pineapple Express, where the air is warmer and that warm air is being transported into Utah? Or is it more complicated.

    1. There is a strong contrast in temperature across the system, so small changes in position can make a big difference on snow level. This is one of several reasons why one needs to be careful about pinpointing snow level so far in advance. In addition, cooling from melting snow varies with precipitation intensity, and that is also hard to nail down.

      Airmass origin is a major reason for the warmth. There are probably some additional effects (heating due to condensation over upstream ranges, etc.), but that's the main one.


  2. In short, fingers crossed, depending on which resort you work/play at . . .