When discussing the state of the atmosphere, meteorologists often talk about “500-millibar heights.” Why is this level chosen?
A There are several scientific reasons why 500 mb is a popular surface for showing weather conditions, but I think history has a lot to do with it as well. First, a little background.
The pressure at sea level is roughly 1000 millibars (mb). Pressure decreases with altitude. As you climb a mountain, the pressure is constantly decreasing, and it's harder to breathe because there's less air. The altitude where the pressure is 500 mb is usually between 16,000 and 19,000 feet above sea level. This altitude is called simply the “500-mb height.” The 500-mb pressure surface is not level. Rather, it is undulating, and meteorologists draw height contours to show how the altitude of this surface changes from place to place.
Even a cursory examination of the 500-mb height contours—these follow the wind—over a few days reveals traveling waves, which, at mid-latitudes, generally move from west to east. In the Northern Hemisphere, the wave crests are called ridges, and the valleys between ridges are called troughs. Troughs tend to bring disturbed and generally cooler weather; ridges bring benign and generally warmer weather.
Here are some scientific reasons why weather conditions are often depicted on the 500-mb surface:
1. About half the mass of the atmosphere lies below 500 mb and half above. In addition, the 500-mb surface is about halfway between sea level and the top of the troposphere—the atmospheric layer where clouds and precipitation occur.
2. The wind closely follows contours of height on the 500-mb pressure surface. This is not true near the ground, where the rough surface retards the wind, causing it to spiral inward toward lower heights and outward from greater heights on a constant pressure surface. To say this another way, the wind crosses the height contours on a constant pressure surface near the ground, but tends to parallel the height contours at 500 mb.
3. The 500-mb surface is high enough that the typical daily temperature cycle (cool in the early morning, warm in the afternoon) is hardly noticeable. Why? Because the major heat input during the day occurs at the ground from the absorption of solar energy. This heat is transferred to the atmosphere by conduction and convection (thermals), but the higher one goes in the troposphere, the smaller the temperature changes between day and night.
4. The flow pattern at 500 mb closely resembles that of the jet stream, a ribbon of strong winds that encircles the mid-latitudes where commercial jets fly, roughly between the 300- and 190-mb pressure surfaces, equivalent to between 30,000 and 40,000 feet. The jet stream is stronger and centered at lower latitudes in winter than in summer.
5. Finally, the 500-mb pressure surface is usually close to the level of nondivergence. What is divergence? It is a property of the air flow. Figure 1 shows an example of pure divergence in two dimensions. The arrows represent wind direction. Air is flowing outward from a common center. In atmospheric flows, pure divergence is seldom realized, but some divergence is usually present and can be computed from the eastward and northward components of the wind. The opposite of divergence is convergence (Figure 2), when air is flowing inward toward a common center. Pure convergence is rarely observed, but you can often see convergence around the periphery of a low-pressure center, where winds cross the pressure contours as they spiral in toward lower pressure. On a constant pressure surface near the ground, you would see the equivalent process: winds spiraling inward across height contours toward lower heights.
Figure 1. Pure divergence in two dimensions: radial winds flowing outward from a common center.
Let's digress a bit and discuss an idealized (and admittedly simplistic) model of a low-pressure system at the surface. Air typically spirals inward toward the center of the low. It also converges along the wind shift lines that define cold and warm fronts. What happens to the converging air? Does it simply pile up at the center of low pressure or along fronts? No. The only place it can go is up. Low-level convergence explains the rising air along fronts and in the vicinity of low-pressure systems. The rising air cools, vapor condenses into clouds, and precipitation often falls.
The air doesn't rise forever because it eventually encounters an inversion, a stable layer at the top of the troposphere, which retards further ascent. This is the tropopause, a mostly impermeable lid on vertical motion. The rising masses of air in a low-pressure environment must go somewhere, so they spread out laterally underneath the tropopause—an event that goes hand in hand with upper tropospheric divergence.
The level of nondivergence separates the lower troposphere, where convergence forces upward motion, from the upper troposphere, where the stable tropopause layer puts a brake on upward motion and forces divergence. Statistically speaking, this level is often close to 500 mb for large-scale low- and high-pressure systems.
Figure 2. Pure convergence in two dimensions: radial winds flowing inward toward a common center; the opposite of divergence.
Figure 3 illustrates the above argument. The vertical axis at the center of the diagram gives pressure in millibars (mb), with pressure decreasing upward and height increasing upward. The horizontal axis is used for two different quantities: divergence, δ, which has units of “per second” (s—1), and vertical velocity, Ω, which has units of millibars per second (mb s−1). We usually think of vertical velocity in terms of feet per second, positive upward, but meteorologists also think of it in terms of millibars per second, which is negative upward. For example, a volume of air whose pressure changes from 850 to 700 mb is definitely rising but its change in pressure over time is negative.
The zero at the bottom of the pressure axis refers to both δ and ω. Convergence (δ < 0) near the ground forces upward motion (ω < 0). Rising motion increases (ω becomes more negative) as altitude increases, but a transition occurs at the level of nondivergence (LND), where δ becomes positive. ω is most negative (rising motion greatest) at the LND, but then begins to decrease as soon as divergence takes over (δ > 0). At a pressure of 200 mb, near the tropopause, vertical velocity (and ω) both become nearly zero.
The LND has historical as well as scientific significance. As the Glossary of Meteorology states, assuming the existence of LND “in theoretical work facilitated the construction of early models in numerical forecasting,” which enabled the first weather predictions by computer. These primitive forecasts were of the 500-mb height contours. This brings us to a discussion of additional historical reasons why the 500-mb surface is so popular.
Since 1909, Weather Bureau meteorologists had been launching pilot balloons with a known rate of rise, and tracking their azimuth and elevation angles with a device called a theodolite. This made it possible to determine winds at various altitudes. But bad weather prevented balloon launches, and balloons often disappeared into low clouds. Figure 4 shows United States locations capable of launching pilot balloons in 1933.
Even before 1900, kites had carried aloft meteographs that recorded ambient temperature, humidity, and pressure. Some of these kites were very large, the size of a subcompact car. However, no wind meant no kite flying. Launching, flying, and reeling in the kites was expensive, labor-intensive, and time-consuming.
From the mid-1920s to the early 1930s, aircraft carried meteographs aloft, but analysis of the measurements had to wait until the aircraft landed, and bad weather occasionally prevented aircraft from flying.
The height limit for pilot balloons, kites, and aircraft alike seldom exceeded 16,000 feet, which doesn't quite reach the 500-mb surface.
In 1933, a Science Advisory Board, serving President Roosevelt, issued a report that, among other things, emphasized the great value to the nation of Weather Bureau operations and made this cogent observation:
During the last decade there has been very rapid progress in Europe in the development and general use of air-mass analysis methods. These require a knowledge of temperatures, humidities, and pressures aloft as well as on the surface, but thus far no systematic attempt has been made to obtain at a given time upper air measurements of these aerological conditions at a considerable number of stations scattered systematically throughout the country so as to make possible the drawing of a daily upper air map of the whole country similar to the surface maps now provided by the Weather Bureau.
Figure 4. Pilot Balloon Stations in 1933.
The board recommended that the number of locations taking routine upper-air observations in the United States be increased to 20 or 25. It also recommended that Willis Gregg, who then headed the Weather Bureau's Aerological Division, be appointed its chief, with the expectation that he would enhance the existing upper air network. Gregg became the Weather Bureau chief on January 31, 1934, and he met expectations. For several years scientists had been testing radio transmitters carried aloft by balloons, realizing that weather data transmitted in real time to the ground would be far more valuable than data recovered many hours after acquisition. In 1937, the year before Gregg died in office, the Weather Bureau inaugurated the radiosonde program, whereby meteograph data were radioed to the ground from altitudes reaching the stratosphere. Radiosonde technology has advanced many times since then. Radiosonde balloons are still being launched today; they now carry a global positioning system (GPS) that gives precise 3-D locations.
By 1940, enough radiosondes were measuring tropospheric conditions to permit regular analyses of conditions at 500 mb, but the earliest published chart I could find was in our NOAA library: the Daily Series Synoptic Weather Maps, Northern Hemisphere Sea Level and 500-Millibar Charts for January 1945 (U.S. Department of Commerce, Weather Bureau). The charts were plotted and analyzed by hand once a day and were valid at 0400 GMT (Greenwich Mean Time). More data were apparently available at 0400 GMT than at other times. Observations within three hours either side of 0400 GMT were plotted, as were wind data near 19,000-foot altitudes.
The National Weather Service and predecessor organizations have been producing The Daily Weather Map of surface conditions for the lower 48 states since 1871. You can find these maps at http://www.lib.noaa.gov/collections/imgdocmaps/daily_weather_maps.html. There have been many changes in the types of information presented and the format, but two are relevant here. The first upper air chart—for 700 mb and for North America—appeared on July 1, 1948. The 500-mb chart replaced the 700-mb chart on May 14, 1954, and has been part of the daily archive ever since.
It is interesting to compare the first archived 500-mb chart with those that NWS forecasters have available today. Figure 5 is the 500-mb chart for 11:00 p.m. Eastern Standard Time, May 13, 1954. It contains plotted wind data at several dozen upper air sounding stations, contours giving the height of the 500-mb pressure surface at 200-foot intervals, and temperature contours at intervals of 5°C. This chart was prepared after the fact for the archives, but I believe that forecasters had to plot and analyze such charts by hand in real time from data received on teletype machines. The continuous feed facsimile charts on a roll of crinkly paper about two and a half feet wide, centrally produced, did not arrive until 1955 and later. Forecasters hung these charts on the walls.
Figure 5. The first 500-mb chart to appear in the Daily Weather Map series, valid at 11:00 p.m. Eastern Standard Time, May 13, 1954. Plotted winds appear at each radiosonde station. Small barbs on the wind staff represent 5 knots, large barbs 10 knots, and flags 50 knots. Solid contours give the height of the 500-mb surface above sea level at 200-ft intervals. Dashed contours give the temperature at 5°C intervals.
Figure 6. A modern 500-mb chart for 1200 GMT July 11, 2016. The station circles locate radiosonde launch points. Temperature (°C) appears to the upper left of the station circle. Dewpoint depression (°C – difference between temperature and dewpoint) is at lower left. The height of the 500-mb surface in tens of meters is at upper right. Example: 594 at Key West, Florida stands for 5,940 meters. The plotting convention for wind is the same as for Figure 5. The station circle is darkened wherever the dewpoint depression is 5°C or less. Height contours are drawn at 30-meter intervals. Stars give the location of cloud-drift winds; squares (only a few) give the location of aircraft reports. The numbers by the stars and squares give the altitude of the report in hundreds of feet. Example: 160 near St. Louis means 16,000 ft.
Figure 6 shows a modern 500-mb chart that forecasters can view on their workstations. It is valid for 1200 GMT on July 11, 2016. Data come from three sources:
1. Radiosondes: Balloons carrying an instrument package are released at 1100 and 2300 GMT and transmit measurements of wind, temperature, dewpoint, and pressure. These are used to analyze conditions on constant pressure surfaces, including 500 mb. The analyses are valid one hour after balloon release.
2. Cloud-drift winds: Animation of satellite cloud images shows cloud motion. Infrared sensors can measure cloud-top temperature, and from that, the cloud height can be estimated. The cloud displacement between successive images provides a wind speed and direction.
3. Aircraft reports: Many commercial airlines provide automated reports of wind and temperature, thousands per day at flight altitude, usually above 25,000 feet, but at least a few at 500 mb, when aircraft climb and descend through this pressure surface.
Note that many more data are collected at 500 mb today than in 1954 (Figure 5). There are about 70 radiosonde stations (circles on the map) in the Lower 48 states today as opposed to a few dozen in 1954. Aircraft reports (squares) and cloud-drift winds (stars) were not available until much more recently.
As hinted earlier, events at 500 mb and surface weather are closely related. You can find an excellent tutorial that explains this relationship in the Mariners Weather Log for December 2008: http://www.vos.noaa.gov/MWL/dec_08/milibar_chart.shtml.
Weatherwise Contributing Editor THOMAS W. 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, 530 Walnut Street, Suite 850, Philadelphia, PA 19106.
I thank Sean Potter, weather historian and Weatherwise contributing editor, for steering me toward a number of references essential for writing this column.