Q: I often hear meteorologists refer to an approaching low pressure system as an “upper low.” Can you describe an upper low and its connection to surface weather?
A: An “upper low” (also known as an upper-level cyclone) is a cyclonic circulation (counter-clockwise in the Northern Hemisphere) above the surface, sometimes pushed along slowly by the prevailing westerlies, but sometimes stalled in one place and disconnected from the westerlies. Upper lows often form when a wave in the westerlies amplifies and evolves from a dip in the westerlies to a closed circulation. A surface low is often associated with the upper low, and the former usually intensifies as the upper wave amplifies. However, if the upper wave forms a closed circulation aloft, the surface low usually weakens soon after. Cool, unsettled, and showery weather often accompanies upper lows.
Here are the details. Figure 1 shows the evolution of an upper low. It is taken from a tutorial that is available at http://www.zamg.ac.at/docu/Manual/SatManu/main.htm?/docu/Manual/. Click on “Conceptual Models,” then on “Upper Level Low.” The sources of the tutorial are the Central Institute for Meteorology and Geodynamics in Austria and the Finnish Meteorological Institute. In each panel, the blue lines indicate the wind flow in the middle or upper troposphere; the red dashed lines indicate temperature contours, with colder air at the top of the image and warmer air at the bottom.
In Panel A of Figure 1, the flow is basically from west to east, with a small dip or trough in the westerlies at the center of the diagram. Colder air lags behind the trough. This is evident if you follow a particular flow line upstream from right to left. The temperature drops as you approach the trough, reaches a minimum behind the trough, and then rises again toward the left edge of the image. A weak surface low-pressure system precedes the trough aloft.
Panel B shows the trough at a later time and a downstream location. By now the trough is amplifying (growing more pronounced). The coldest air still lags behind the trough. The surface low is intensifying, with increasing wind and precipitation.
In Panel C, the trough has slowed its eastward progression considerably, and cold air has entered the base of the trough so that the wind flow and the temperature contours are parallel. The surface low has reached maximum intensity.
In Panel D, the upper low has developed a closed circulation, centered on a pool of cold air aloft. Because this circulation has separated from the main westerlies near the top of the figure, it is often called a “cut-off low.” By now, the surface low is directly beneath the upper low and is producing gloomy, showery, and cool weather. Such weather may be prolonged because cut-off lows move slowly, if at all. They are often associated with a blocking weather pattern, in which the prevailing westerlies are forced northward or southward around the cut-off low.
Caption: Figure 1. The evolution of a traveling wave in westerly flow into an upper-air or cut-off low. Each panel covers an area roughly 1,200 miles (left to right) by 900 miles (top to bottom). Red dashed lines are temperature contours with the coldest air near the top of each diagram. Blue arrows show the direction of flow in the middle or upper troposphere. Closer spacing implies greater wind speed. A minor dip in the westerlies at the center of (A) is followed by colder air (the dip in the temperature contours). In (B) the trough in the westerlies and in the temperature contours amplifies. In (C), the cold air enters the base of the trough, and a closed circulation begins to form. In (D), the upper-air low has become cut off from the main westerly flow and sits, spinning in a pool of cold air aloft.
Caption: Figure 2. Same as Figure 1, but illustrating the demise of a cut-off low through capture by a traveling wave in the westerlies. (A) The cut-off low is isolated from the jet stream to the north. (B) A trough in the westerlies approaches the cut-off low. (C) The trough captures the cut-off low and sweeps it northeastward while weakening its circulation.
Cut-off lows can form anywhere in the lower 48 states, but they are most prevalent near southern California or just off the coast during winter and spring.
What happens to upper lows once formed? They may sit and spin for days, finally weakening to the point where they are no longer recognizable. Or they may be picked up by an approaching trough in the westerlies, in which case the closed circulation quickly disappears and they are carried away as a minor trough. Figure 2 illustrates the latter case. In Panel A, the cut-off low is completely isolated from the westerlies. In Panel B, a trough approaches at a latitude sufficiently far south to capture the cut-off low. This happens in Panel C as the cut-off low is swept up by the jet stream and carried northeast as a short-wave trough.
Upper lows should be distinguished from “upper waves” or “upper troughs.” The latter usually move along in the westerlies at a good clip. They can cause stormy weather, which is usually short-lived because they move much more rapidly than upper lows.
Q: On January 21, 2011, an intensifying wave of low pressure passed near Cape Cod, Massachusetts, delivering about 10 hours of light-to-moderate snow with generally light wind. At 8:30 a.m., near the midpoint of the snowfall, I measured 3.2 inches on a white car top and brushed it away. I brought my precipitation gauge inside and immediately replaced it with a second gauge. I melted the snow in the first gauge; the water equivalent depth was 0.32 inches. At 12:45 p.m., I measured a further accumulation of 3.0 inches for a total depth of 6.2 inches. The snow in the replacement gauge melted to 0.15 inches. I've included photos of my precipitation gauges just before measurement. How do I get the snow-to-water ratio for the entire amount of snow, and why did the ratio change so much during the storm? What is the standard time interval for measuring snowfall? Because less snow is captured in the gauge when it's windy, under what conditions is it inadvisable to use a gauge to get an accurate measure of the water content in snow?
A: Measuring snow depth can get complicated especially if the snow (1) is intermittent, (2) is mixed with rain or sleet, (3) is falling on warm ground or when the surface temperature is near or above freezing, or (4) is accompanied by strong winds. For brevity, I'll restrict comments to the snowfall you describe.
For the first period of snow, the snow-to-water ratio is 3.2/0.32 or 10 to 1. For the second period, it is 3.0/0.15 or 20 to 1. For the entire storm, the ratio is 6.2/0.47 or 13.2 to 1. One could compute the storm-average ratio by accounting for the time intervals associated with each measurement, but this is not normally done.
Your photos suggest, and your measurements confirm, roughly the same snow depths for each period, but the density of the fallen snow clearly changed between the beginning and end of the storm, which probably means the crystal type changed as well. Often, the temperature in snow-bearing clouds decreases during an extended storm. Both the temperature and the vapor supply within the cloud determine the crystal type, and so the latter will change during a snowfall when in-cloud conditions change. Since you reported a 10:1 ratio in the first part of the storm, it is likely that plate, column, or needle crystals were falling, characteristic of in-cloud temperatures in the range 0°C to –10°C. The 20:1 ratio reported during the latter part of the storm suggests a fall of spatial dendrites: large, stellar crystals with delicate shapes. These are favored when the in-cloud temperature is around –15°C and the vapor supply is rich. These crystals, and larger flakes formed when many of them stick together, result in low-density snow. In Colorado, where this is common, the ratio of snow depth to liquid water equivalent is often 20:1 or even 30:1 in fresh snow that hasn't yet settled. If you want to know more about how in-cloud temperature and vapor supply determine snow crystal type, go to http://www.its.caltech.edu/~atomic/snowcrystals/primer/morphologydiagram.jpg.
The National Weather Service (NWS) suggests that snow depth not be measured more often than once every six hours and that no more than four measurements be made during a 24-hour period. A “snow board,” e.g., a two-foot square piece of plywood painted white, is very handy for making repeated snow-depth measurements. The board should be swept clean after each measurement and placed so that it is even with the top of the current snow surface. If the snow is fluffy, you may need to prop the board up at the corners. Failure to do so may result in drifting snow covering the board and a later measurement of snow that did not fall from the sky. The storm total snowfall is the sum of all measurements made during the storm. It will almost always differ from the maximum snow depth on the ground, because snow, especially fluffy snow, settles as it falls, and bottom melting can occur if the snow falls on warm ground. Following exceptional storms, observers often debate whether any records were set. Those who sweep and measure every hour will tally greater snowfall than those who measure every six hours during the same storm. The NWS wants to standardize measurements, thus its recommendation that they be spaced at least six hours apart.
Caption: Four-inch-diameter precipitation gauge used to measure water-equivalent depth of snow. Top: Photographed at 8:30 a.m. EST January 21, 2011, just before removal to indoors. Bottom: This gauge was placed outside at 8:30 a.m. and photographed at 1:00 p.m. just before removal to indoors.
Snowfall measurements become notoriously difficult when it's windy. Undercatch of precipitation is common when wind-induced updrafts above the orifice of the gauge carry snowflakes over the collecting cylinder when they would otherwise fall in. An article by Roy Rasmussen and multiple coauthors in the June 2012 Monthly Weather Review called “How Well Are We Measuring Snow?” reveals how serious this problem is: the undercatch can range from 20-50 percent of the correct amount. Many elaborate wind shields have been invented to reduce the undercatch. The designated standard for snowfall measurement in windy conditions is a double wind fence, octagonal in shape. The diameters of the outer and inner fences, constructed of wooden slats with open spaces in between them, are 39 and 13 feet, respectively. The fences are elevated above the surface so as not to impede drifting snow.
Such elaborate wind shields are beyond the reach of individual observers, and so the best advice for measuring snow during and immediately after a windy storm is to find a flat, open area, away from obstacles, where drifting is minimized. Stick the ruler into the snow at several locations and compute the average depth. If you have a standard, cylindrical, clear-plastic gauge, it will not have collected a representative amount of snow, but you can empty the gauge, invert it in the same open area where you measure snow depth, and push it all the way to the ground. Slide a piece of cardboard underneath the gauge to hold the snow in, then turn it right-side up again. Melt what is in the cylinder to get the water equivalent. For hydrological purposes, in particular, making estimates of future runoff, there's no substitute for obtaining a vertical core sample in a tube and melting or weighing it to obtain the water equivalent.
The above technique for measuring snow and many others are thoroughly illustrated with photos at http://www.cocorahs.org/media/docs/Measuring%20Snow-National-Training%201.1.pdf.
Q: I used to live near a lake and now live in Lubec, Maine, near the ocean. We get sea smoke here—at least that's what people call it. Occasional wisps of fog used to appear over the lake where we lived, but it's nothing like this sea smoke here in Maine. What causes it? The photo [see below] was taken on January 16, 2012, when the air temperature was about 12°F.
A: In late fall and early winter, before water near the New England Coast freezes, a mass of cold air will sometimes sweep southeast from Canada over much warmer water. Sometimes the temperature contrast between the water surface and the overlying air can be extreme, 30°F or more. Under such conditions, the amount of water vapor and sensible heat transferred from water to air becomes quite large. I'll describe these transfers separately.
Caption: Sea smoke on a cold morning, January 16, 2012, looking east from West Quoddy Head, Lubec, Maine. A scallop dragger is at center. Grand Manan Island, New Brunswick, is in the distance.
The vapor pressure at the water surface depends only on its temperature. For example, if the water surface is 5°C (41°F), the vapor pressure at the surface is 8.7 millibars. If the temperature of the overlying air is –10°C (14°F), the saturation vapor pressure (the maximum partial pressure that water vapor can have without condensation) is 2.9 millibars. This creates a strong gradient in vapor pressure just above the water surface. The vapor transfer from water surface to air is so rapid that some of it condenses in the cold air, causing sea smoke.
Sensible heat transfer upward from the surface is rapid, too. The cold air in contact with the much warmer water is heated from below, creating a steep lapse rate (rapid decrease in temperature with altitude). The usual maximum lapse rate observed is about 1°C of cooling for every 100-meter increase in altitude, but within a meter or so of the water surface, it can be higher. This causes the sea smoke to twist and curl upward, much like smoke from a fire—hence the name.
With offshore flow, sea smoke may be the only manifestation of cold air passing over much warmer water close to the coast, but, with increasing distance from the coast, the layer of air with a steep lapse rate grows rapidly upward, and low stratocumulus clouds form and thicken. Still farther offshore, snow showers form. Lake-effect snows develop toward the downwind shore of the Great Lakes every winter from the same cause.
When exceptionally cold air sweeps off the mid-Atlantic coast and over the Gulf Stream, seafarers occasionally witness a chaotic sight: sea smoke rising from the surface all the way into turbulent low clouds and vortices of condensed vapor extending upward from the surface in slate-gray columns.
Q: A friend of mine sent me some photos of Niagara Falls almost completely frozen over [see example, below], presumably sometime in 1911. These photos have been making the rounds on the Web. What weather conditions back then could have caused the falls to freeze over?
East Bethel, Minnesota
A: We've reprinted one of the photos you sent, mainly because it is accompanied by a date: 1911. Snopes.com, founded in 1995 by Barbara and David Mikkelson to investigate urban legends and popular rumors, is considered an authoritative source. You can find the Snopes discussion of all four photos commonly circulated at http://www.snopes.com/photos/natural/niagarafalls.asp. Snopes believes the four photos were not all taken on the same dates. It considers the photo printed here to be unretouched, but the date is uncertain. Intrigued by the information on this Web site, I contacted David Zaff, a meteorologist at the Buffalo, New York, office of the NWS. (Buffalo is at the east end of Lake Erie, which drains into the Niagara River. Niagara Falls is just a few miles downstream from Buffalo.) David, in turn, pointed me to the Web site of the Northeast Regional Climate Center, http://xmacis.rcc-acis.org/. From there, I was able to generate a table of the coldest months ever experienced at Buffalo since records began in 1873 (see Table 1).
Caption: Photo of a nearly frozen Niagara Falls, purportedly taken in 1911, but more likely in January 1912.
Table: Table 1. Lowest Mean Monthly Temperatures Ever Recorded at Buffalo, New York, from August 1873 Through May 2013
January 1912 was a very cold month in Buffalo. A New York Times article, mentioned by Snopes and printed February 5, 1912, mentions a large ice bridge above the falls that had been in place for a week, building to a height of 60-80 feet as ice floating down the river piled up behind it. It was the third ice bridge to have formed during the 1911-1912 winter. The main thrust of the article was a description of the tragic deaths of three people who were swept over the falls when the ice bridge unexpectedly gave way on February 4, a “severely cold day.” In the week before, many adventurous souls had ventured onto the ice without incident.
Based on the foregoing, if the photo was indeed taken during the 1911-1912 winter, it was probably taken in January 1912. Careful inspection reveals that a few small streams of water are still flowing over the precipice, contrary to the claim that the falls were “completely frozen.”
Weatherwise Contributing Editor THOMAS SCHLATTER is a retired meteorologist and volunteer at NOAA's Earth System Research Laboratory in Boulder, Colorado. Submit queries to the author at firstname.lastname@example.org, or by mail in care of Weatherwise, Taylor & Francis, 325 Chestnut St., Suite 800, Philadelphia, PA 19106.