AD Header Dropdowns

AD Main Menu

Despite record snows, Anchorage's breakup smooth so far

John PapineauNational Weather Service
Charlie Earnshaw photo

After Anchorage's record snowfall (134.5 inches in West Anchorage to 220 inches on the upper Hillside) there was considerable trepidation about how severe our "breakup" would be this year. Where, many wondered, would all that water go?

First, some definitions. In a strict sense, "breakup" refers to the breaking up of river ice and its subsequent movement, as is common on the Kuskokwim and Yukon Rivers in May. If the ice remains in place once snowmelt runoff enters the rivers, or if the ice moves and then congregates to form a jam, flooding ensues. Anchorage, of course, does not experience this type of breakup. Ours is simple melting and snowmelt runoff in an urban setting.

So what controls the severity or amount of water available to pond on Anchorage streets and low-lying areas? Generally, our roads are designed to shed water, whether snow melt or rain.  Roads themselves are convex so water runs off the sides, and in many places there are ditches off the sides that store or transport water to ponds or creeks. Critical is the speed of the melt, especially the first week or two. To appreciate how this works, some understanding of the snowmelt process helps.

Air temperature and sunlight

A pile of snow melts top down, as the presence of frozen ground restricts heat transfer into the snow from the upper layers of soil. However, keep in mind that for most of the season the lower section of a deep snow pack is considerably warmer than the upper layer, which fluctuates with air temperatures. Snow melts for short periods all winter, but the main event starts in late March and continues through April. The two driving controls of snowmelt are air temperature and sunlight, which we'll consider separately.

When the air directly above the surface is warmer than the snow, heat is transferred to the snow, which warms the snow crystals. However, the temperature of the snow crystals can never exceed the freezing point (32 degrees F). Excess heat is used to melt the crystals. If enough heat is transferred to the snow, a layer of water may form near the surface.

What happens next? Some water evaporates, but most of it percolates through the snowpack to deeper layers. At some point the water reaches a layer where the snow is colder than 32 degrees, and the water starts to re-freeze. But in so doing, water gives up heat to the snow (latent heat exchange due to phase change).  

This is an ongoing process during the first phase of the spring snowmelt season, something the snow geeks call "ripening." Any ripening is complicated by another factor. At night, when the temperatures drop below freezing, a "cold wave" works its way through upper layers of the snow (top-down) and causes the water to refreeze and snow crystals may cool below 32 degrees.

Sunlight's role 

So what role does sunlight play? When sunlight strikes the snow surface, several possibilities occur:

  • Most is reflected back to the air if the snow is clean, so only a small portion of sunlight is available to melt snow.  
  • When snow is dirty, a large portion of the sunlight is absorbed and used to melt snow crystals. Sunlight can penetrate several inches and produce melt layers below the surface, giving the upper layer a hollow structure.  
  • Snow itself is a good absorber and emitter of infrared radiation, which means that it's constantly attempting to reach some type of thermal balance. When the energy -- all forms -- added to the snow exceeds the energy that is given off, heavy-duty melting begin. 

An additional factor controlling the severity of the meltdown is how much water percolates into the ground. Recall that in late autumn, as air temperatures fall below freezing, the ground starts to freeze. If soil contains much moisture, it forms a layer of frozen ground relatively impervious to water. The depth of this frozen material, what we refer to as the frost depth, is a function of the air temperatures and amount of snowfall. A deep snow pack that forms early in the season, as it did this year, insulates the ground and limits the frost depth. What all this means in the spring is that during the first phase of meltdown (lasting up to 10 days) little water is stored in the ground. It runs off the soil and other impermeable surfaces such as concrete and asphalt. Later, however, after the first few inches of frozen ground melts, some snowmelt is stored in this layer or "flows" through the layer to low-lying areas.

Some snow loss is due to evaporation from the snow surface. And some snow crystals sublimate, too -- that is, change directly from a solid to vapor. The loss of water due to these two processes during the spring is probably small compared to the overall amount of water lost each day.

Rain and Chinooks

Two other physical processes can melt snow in a hurry, but typically do not occur in the spring -- rain-on-snow events and Chinooks.  

Rain-on-snow is pretty rare, although it did occur Dec. 11, 2011 -- especially on the Hillside -- during one of our down-sloping wind events. Rainwater is, of course, above freezing, although sometimes just barely. Nevertheless, it delivers heat to the snowpack. A warm rain can melt a lot of snow quickly, and is a major flood concern in the Pacific Northwest.  

The other phenomenon that occurs frequently during the heart of winter -- less so in March and April -- are down-sloping windstorms or Chinooks, which are sometimes called "snow eaters." Warm dry air constantly moves over the snow surface, which produces melting, then rapid evaporation. As a result, most Chinooks produce less melt water runoff than one would expect otherwise given the loss of snow, but can remove inches of water equivalent in a day or two.

So what dictates whether Anchorage is going to have snowmelt problems?

  • Melt speed. Rapid melting means a greater chance of water ponding on roadways. This is especially true during the beginning of the snowmelt, when the ground is still frozen. The speed of melting in turn is dictated by how much sunlight reaches the snowpack and the warmth of the air.  
  • Cloud cover slows the melt, while a few days of cooler temperatures can also put the brakes on. Keep in mind that rapid snowmelt is a relative term. At the end of March, the National Weather Service forecast office had 26 inches of snow on the ground, with a water equivalent of about 6.4 inches. That means if the 26 inches melted instantly, it would turn into 6.4 inches of water. But a typical snowmelt season lasts 25-30 days. Hence we lose about one-quarter to one-third of an inch of water equivalent per day.  

As April progresses with steadily warming temperatures, snowmelt accelerates if there are no major changes in weather patterns.

Painless breakup

This year, despite slightly warmer-than-average April temperatures and large amounts of snow, the meltdown has been relatively painless.

The warm-up in early April helped make breakup gradual (Figure 1). Slow melting of the snowpack avoided a rapid flush of water that might have led to flooding. In contrast, look at the last several days of March in 1995 (Anchorage's fourth snowiest season), which saw a rapid increase in temperatures with plenty of sun (Figure 2). There was a similar amount of snow on the ground at that time, and many readers will recall a number of minor floods around town.

If you have piles of snow you want to see melt quickly, sprinkle some dark-colored environmentally friendly substance onto the snow, like ash from the fireplace. This allows more absorption by sunlight. Better yet, if you can still manage a snow shovel, spread the snow across the ground. This enhances snowmelt because you have increased the snow's surface area  significantly, allowing the warm air to do its work on larger portion of the snowpack.  

And all this time you thought snowmelt was child's play.

John Papineau is a meteorologist with the National Weather Service in Anchorage. Reach him at john.papineau@noaa.gov. Used with permission.