Clouds in the Arctic:     A Tale of Diamond Dust and Summertime Stratus.

a webpage for the ATOC 5600 final by Dave Porter

TOC

    Clouds in the Arctic

    Summertime Stratus

    Diamond Dust

    References

Cloud cover is a very important factor that influences the global radiation budget.  The arctic is a very cloudy place.  This is especially true in summer when low-lying stratus if prevalent.  Here, I will discuss research and theories on the most important cloud-cover type in the Arctic, summertime stratus, as well as a unique phenomenon known as "diamond dust".
 Clouds in the Arctic   
The most important characteristic of clouds in the Arctic is the summer stratus. From about mid-June to mid-September, the ocean area covered by sea ice is 80 to 90 percent covered with this cloud type. Summer stratus has important effects on the radiation balance of the surface.  The effect of clouds in the arctic is to reduce the downwelling solar flux at the surface.  This is mainly due to the high albedo, 60-75%, for arctic stratus (Herman and Curry, 1984).  This attenuation of the downward shortwave flux is partially modified by a subsequent increase in downwelling shortwave radiation at the surface by multiple reflections between the surface and cloud base (Wendler et al., 1981).

The most prominent feature of low cloud cover during the winter months is the maximum over the northern North Atlantic. This is a reflection of the uplift of air masses by the frequent cyclone activity in this area associated with the North Atlantic Stormtrack. Low cloud cover is rather limited over central and eastern Siberia because of a lack of moisture and the general subsidence of air in the area of the strong Siberian high.

During summer, the Icelandic Low and Siberian High weaken. Low cloud cover becomes more uniform, but with a distinct increase over the Arctic Ocean. The increase reflects the dominance of low-level stratus, which form as warm air masses moving over the ocean are chilled by the cold, melting sea ice cover.

Then autumn months illustrate the transition back to the winter pattern. Total cloud cover combines low, middle and high clouds. While amounts of total cloud are hence greater than for only low cloud, it can be seen that most cloud cover is of the low variety.
 TOP
Summertime Stratus
  
Global climate models have particular difficulty in simulating the low-level clouds during the Arctic summer.  Model problems are exacerbated in the polar regions by the complicated vertical structure of the Arctic boundary layer.  The presence of multiple cloud layers, a humidity inversion above cloud top, and vertical fluxes in the cloud that are decoupled from the surface fluxes, identified in Curry et al. (1988), suggest that models containing sophisticated physical parameterizations would be required to accurately model this region.  Accurate modeling of the vertical structure of multiple cloud layers in climate models is important for determination of the surface radiative fluxes.

As seen in the figure to the left, there is a pronounced maximum of low level clouds in the Arctic basin in the summer months.



Arctic low-level stratus, especially in summer, when melting ice pack creates a uniform lower temperature at 0 degC over which warm air from lower latitudes moves.  As this warm air is cooled, condensation is formed in the lower boundary layer due to adiabatic ascent.  Because the sun remains low in the sky, even in the summer months, stratus clouds evolve.  The monthly mean cloud amount is over 80% during the summer and are dominated by low-level cloud  types (Houze).

(Figure 1 - Frequency of various cloud types and precipitation from Huschke, 1969)


In the arctic, cloud models include a mean downward vertical velocity (~mm/s).  This illustrates how clouds here do not depends on lifting for their exhistence and can thrive in an environment with mean downward motion.

The stratus, which develops after a week, has a decidedly layered structure with a more dense uper and more tenuous lower layer.



(Figure 2 - Radiative and turbulent processes producing the structure of summertime stratus from Herman and Goody, 1976)

As the warm air moves over the surface boundary at 0 degC, the turbulent mixing rapidly spreads the cooling upward by both latent and sensible fluxes.  In one days time, the cooling and IR cooling leads to condensation, leading to cloud formation.  The absorptive properties of the droplets significantly changes the radiative properties of the atmospheric layer.  The upper and more dense cloud layer becomes unstable by longwave cooling above and IR warming from below.  The upper cloud retains a steady-state character at the end of 7 days.  The lower and more tenuous layer of the cloud is balanced by a balance between the insolation penetrating into it from above and cooling from the melting ice below.  The middle, clear layer, is in a state of purely radiative equilibrium.  In the simulation used to generate the above diagram, radiation fog just above the surface during the very first hour is missing.  It is this fog that is lifted over the stable boundary layer.  The presence of  fog is determined by the weakenss of the mixing (Houze).  In time, ice precipitation moistens and cools the lower boundary layer and deepen and thermodynamically stabalize their environment.  If ice sedimentation is rapid enough, and the cloud capping of the boundary layer becomes optically thin through ice depositional processes, a lower cloud layer forms through the radiative cooling of a moisture inversion.  Once the condensate forms, it wants to remain since the dissipative mechanisms at mid-latitudes are weak in the arctic.  One of the main reasons of the pronounced summer maximum is because this is the season where normal dissipative processes are not operating.  Convection in the transition seasons and a lack of moisture in the winter prohibit stratus formation.
These clouds have been described by Professor N. Untersteiner, who was present at the ice station at the first IGY for 366 days, as almost continuously present low stratus, of a "boring" character, interrupted only occasionally by breaks in the clouds cover and occasional fog (Houze).

Why are their not persistent summertime  stratus over Greenland or Antarctica?  Is is likely that orography is the dominant factor as katabatic winds would act to de-stable and mix the boundary layer.  Also, there are persistant anti-cyclones in these regions, and a large diurnal insolation cycle over Greenland, that act to inhibit summertime stratus (Herman and Goody, 1976).
 TOP
Diamond Dust
  

Diamond dust is defined in the AMS glossary as, "small ice crystals falling from an apparently cloudless sky, (often, but not always, at night)".  It is reported in METAR codes as IC, for ice crystals.  Crystals originate from air having a higher moisture content above a thermal inversion aloft, where mixing leads to nucleation and growth of crystals at temperatures near −40°C.  It has been shown that diamond dust can be pushed out of the way while being suspended in air, and there are several stories of people walking though the ice crystals, turning around and seeing a tunnel in the shape of an outline of their body.

There is a distinction between freezing fog, which is waer droplets that freeze after condensing, and diamond dust, which is the direct deposition of water vapor into ice crystals.  This process causes mostly uniform size and shape (hexagonal) of the formed ice crystals.  Also, fog acts to significantly reduce visibility, where as diamond dust is optically thin.



While diamond dust can be seen in any area of the world that has cold winters, it is most frequent in the interior of Antarctica, where it is common year-round. Schwerdtfeger (1970) shows that diamond dust was observed on average 316 days a year at Plateau Station in Antarctica, and Radok and Lile (1977) estimate that over 70% of the precipitation that fell at Plateau Station in 1967 fell in the form of diamond dust (once melted, the total precipitation for the year was only 25 mm [1 inch]).

Diamond dust is often associated with halos around the sun and other related optical phenomena (Greenler, 1999). These result because the diamond dust crystals form directly as ice (as opposed to freezing drops), and because they generally form slowly. This combination results in crystals with well defined shapes, usually either hexagonal plates or columns. These shapes, like a prism, can refract light in specific directions. Some halos can also be seen under a cirrus cloud, but diamond dust can create much more spectacular displays because the ice crystals are all around the observer.

The depth of the diamond dust layer can vary substantially from as little as 20 to 30 m (60 to 100 feet) to a few hundred metres (1000 feet). Because diamond dust does not always reduce visibility it is often first noticed by the brief flashes caused when the tiny crystals, tumbling through the air, reflect sunlight to your eye. This glittering effect gives the phenomenon its name since it looks like many tiny diamonds are flashing in the air.


Because of the length of the polar night, clouds are hard to observe for a large portion of the year, and so, diamond dust is significantly over-reported.











(History of ice crystals falling in a slightly ice-supersaturated atmosphere. www.awi.de )

Observations showed that diamond dust contributed only a negligible radiative effect to the sea ice surface. Surface radiative fluxes and radiative forcing values during diamond dust events were similar in magnitude when compared to obser ved clear-sky periods. Combined information from lidar, radar, and surface obser vers showed that diamond dust occurred 13% of the time between November and mid-May over the Arctic Ocean and was not obser ved between mid-May and October. Diamond dust vertical depths, derived from lidar measurements, varied between 100 and 1000 m but were most often observed to be about 250 m.  

Lidar and radar measurements were analyzed to assess if precipitation from boundar y layer clouds was present during times when surface obser vers reported diamond dust. This analysis revealed that surface obser vers had incorrectly coded diamond dust events 45% of the time. The miscoded events occurred almost exclusively under conditions with limited or no illumination (December–March). In 95% of the miscoded reports, lidar measurements revealed the presence of thin liquid water clouds precipitating ice crystals down to the surface (Intrieri and Shupe 2004).



Anthropogenic Formation
When water vapor exits a car tailpipe when it's minus 40, for example, the water vapor temperature drops from about 250 degrees to minus 40 in less than 10 seconds. Water cooled that fast forms tiny ice particles, so small that ten of them could fit side by side on the finger-cutting edge of a piece of paper. Collectively, millions of these particles take form as ice fog, the cotton candy-like clouds that hang over our roads.  In a research paper Benson wrote in 1965, he scouted the sources of water vapor in Fairbanks. He found the cooling water dumped from power plants into local rivers and sloughs accounted for 64 percent of the manmade water sources that could end up as ice fog. Gasoline combustion made up only 3 percent, but that 3 percent hangs where we drive.

Although not nearly the ice-fog machines that vehicles are, humans also are to blame. In his report, Benson calculated people in the Fairbanks area exhaled and sweated about 58 tons of water vapor into the air each day in 1964. Benson even averaged the water output of Fairbanks' dog population. He estimated there were about 2,000 outdoor dogs in Fairbanks in 1964. Those 2,000 resting dogs exhaled a little more than a half ton of water each day into the air, which Benson deemed "a conservative estimate because all of the dogs don't rest all of the time."

Benson figured dogs, open water, cars, stoves and other sources pumped 4,000 tons of water vapor into the air each day in 1964. He estimated the current Fairbanks water output to be perhaps more than 6,000 tons per day (Rozell, www.gi.alaska.edu).
TOP
References 

Clark, M.P., M.C. Serreze, R.G. Barry. 1996. Characteristics of Arctic Ocean Climate Based on COADS Data, 1980-1993. Geophysical Research Letters 23(15):1953-1956.

Curry, J. A., E. E. Ebert, and G. F. Herman.  1988.  Mean and turbulence structure of the summertime Arctic cloudy boundary layer, Quart. J. Roy. Met. Soc., 114, 715-746

Greenler, R. (1999). Rainbows, Halos, and Glories. Milwaukee: Peanut Butter Publishing, 195 pp. ISBN 0897169263.  — An excellent reference for optical phenomena including photos of displays in Antarctica caused by diamond dust. 

House, Robert A., 1993.  Cloud Dynamics. Academic Press.

Intrieri, J., M. Shupe, 2004.  Characteristics and Radiative Effects of Diamond Dust over the Western Arctic ocean region.

Schwerdtfeger, W. (1970). "The climate of the Antarctic", in S. Orvig: Climates of the Polar Regions, World Survey of Climatology. Elsevier, 253-355. ISBN 0444408282.

TOP


By David Porter, 2007