Weather, Fronts, and Forecasts: _{From observations to predictive models}

Introduction

When you are getting ready to go to school or work for the day, where do you get your weather forecast? Maybe you check an app on your phone, listen to the radio, or watch the news on a local TV station. Maybe you have heard your local meteorologist mention a front moving through will bring a drop in temperature and some rain. The weather map shows thick, curved lines with triangles or half-circles, like in Figure 1. But how do meteorologists know where to draw those lines or what fronts are where? How do they know that the fronts will cause the changes in weather that they describe?

Meteorologists today have many sources of data to help them make accurate weather forecasts: thousands of networked ground-based measurements of air temperature, wind speed and direction, and precipitation, as well as widely available near-real-time satellite imagery. These data simplify the identification of current weather patterns. But how does current weather data help them create weather forecasts, predicting what is likely to happen in the future? Addressing that question requires understanding more about the history of weather observations, the scientists who developed models to explain the observations, and how those models inform modern weather forecasting.

Historical weather observations

People around the world have always been keen observers of the weather, often for agricultural and ocean navigation purposes. As societies (or their rulers) sought more standard measures of components of weather like rainfall and wind speed, people invented and refined instruments that could be easily reproduced. Using the same instruments allowed people to collect comparable measurements across a country or region. These instruments included:

• The rain gauge to measure precipitation,
• The hygrometer to measure humidity,
• The anemometer to measure wind speed,
• The wind vane to indicate wind direction,
• The thermometer to measure air temperature, and
• The barometer to measure air pressure.

Versions of these instruments were used in many societies around the world. Some of the earliest records we have today of weather observations come from Britain. By the early 1840s, weather observations were abundantly recorded across the British Isles, but most of these records remained in the hands of local observers and were not standardized across the country.

Around that time, an invention of a different kind allowed observers’ weather observations to be collected and reported for larger regions. In August 1848, the British newspaper The Daily News published the first report of weather conditions across the country, based on observations collected from thirty stations at 9:00 am the previous morning. The observations had been sent to The Daily News by telegraph, the newly developed electrical transmission system that allowed rapid communication over long distances where telegraph wires were in place. The person behind the new observing network was James Glaisher, an avid British meteorologist known for pushing the boundaries of possibility. He had personally inspected the weather stations, wrote detailed directions for when and how to collect the measurements, and then curated the submitted reports and prepared data tables for daily publication.

Glaisher also analyzed the records from the 30 stations, calculating monthly and annual temperature highs, lows, and means; humidity; general wind direction and strength; precipitation; and cloud cover. He compared these data to averages from previous years to describe change over time and published this work annually from 1847 to 1892. Glaisher’s reports were the first of their kind, but are now common practice followed by meteorological offices worldwide.

In 1851, Glaisher pioneered another technique for displaying weather data: hand-drawn maps that allowed readers to see both the data and the distribution of the observations across England (Figure 2). He presented this new technique at the 1851 Great Exhibition in London, where it caught the attention of Prince Albert, the husband of Queen Victoria and an ardent promoter of the exhibition. The Prince invited Glaisher to give a lecture at the Royal Astronomical Society to share his new maps. In his presentation to the society, Glaisher expressed his belief that “a widely spread and universal system of simultaneous observation… must be the groundwork” for establishing meteorology as a science.

Less than a year later, American naval officer Matthew Maury organized an international conference in Brussels, with an agenda that supported Glaisher’s assertion. Maury had developed detailed monthly wind charts for the Atlantic Ocean based on observations recorded in his own and others’ ship logs (see our Ocean Currents module). Ship captains who used the maps to plot their course had reduced their travel times as they traversed the Atlantic. Maury sought to gather more data from other nations’ sailing vessels to create similar maps for the rest of the world’s oceans. Britain and the nine other countries at the conference signed on to this effort, agreeing that the captains of their merchant vessels would collect the same data at the same time of day every day of their voyage: their position, air temperature, pressure, wind direction, and wind speed. This act, in essence, created an ocean-based “widely spread and universal system” that Glaisher envisioned.

In 1854, the British government established the first Meteorological Department to manage all that data, giving oversight to the Board of Trade. Today, it’s known as the Meteorological Office or Met Office. Placing the office in the Board of Trade was a deliberate choice: The primary motivation for collecting data and producing maps was not in the name of science but to increase profits from Britain’s trading industry in the East Indies.

Once the office was established, it needed a director. Glaisher might have seemed like an obvious choice, but he was not interested in the role. The person who lobbied hardest to lead the new office did not consider himself a meteorologist, unlike Glaisher. Instead, like Maury, he brought years of experience as a naval officer making careful weather observations as he navigated the seas. The Board of Trade appointed Robert FitzRoy, famous for serving as the captain of the HMS Beagle that brought Charles Darwin to the Galapagos and South America in the 1830s, as the first director of the Met Office.

Using weather observations for forecasts

Although Glaisher lacked the inclination to take on the Met Office director role, he had already made two critical contributions to the young science of meteorology that helped guide the work of the office. First, he recognized the importance of having a network of observing stations that followed a standard procedure for data collection and shared their data with a central processing hub. Second, he saw the power of maps to transform tables of numeric data into visual data (see our Using Graphs and Visual Data in Science module) and began the now-common practice of plotting weather data on maps that showed towns and geographic markers like coastlines.

As the director of the new Met Office, FitzRoy worked to expand Glaisher’s network of land-based stations across the British Isles. He produced Daily Weather Reports for the newspapers that became more extensive and detailed as the network grew. To support the development of wind maps for the seas, he contacted port administrators and collected weather observations from merchant ship captains when they arrived in port.

But FitzRoy also knew that yesterday’s weather records could not help mariners know when major storms were coming. This, too, was the Met Office's mandate, allowing FitzRoy to delve deeper into the science of meteorology. He was further motivated by a severe storm that hit the British Isles in October, 1859, causing the loss of at least 800 lives and more than 120 ships. The storm became known as the Royal Charter Gale for the clipper ship the Royal Charter, which was carrying passengers from Melbourne, Australia, to England, and wrecked on the Welsh coast. FitzRoy claimed that his office could have provided warnings about the storm's approach through the telegraphic communication of weather reports to ports that would have given them time to prepare and send ships out to sea.

FitzRoy documented his claim by creating a series of weather maps at each hour of the storm, as shown in Figure 3. He called these synoptic maps, meaning they give a general sense of the whole system by combining temperature, pressure, wind, and geography at a large scale. In advance of the storm, the synoptic maps showed a drop in air pressure and temperature, winds that rotated counterclockwise around a calm center, and a whole system that moved along the coast from south to north over several hours. Indeed, with weather reports from the southern coast, they could have posted visual warnings on the Welsh coast that would have kept the Royal Charter out at sea and prevented its wreck.

As a result of his work documenting the storm's progress, FitzRoy was given the go-ahead to develop his Storm Warning Service. He devised a set of cones and drums that could be hoisted like flags up the telegraph poles at ports to indicate the probable direction of an approaching storm to ships up to seven miles out at sea. It relied on the telegraphic transmission of weather data that allowed FitzRoy’s office to see the characteristic drop in pressure and temperature and to relay the warning back to the port. The system was installed at 50 stations in late 1860. In February, 1861, FitzRoy issued the first storm warning, and the system's success in saving ships from wrecking was apparent.

FitzRoy was not content with limiting the service to the marine setting, however. Within a few months, he began adding brief statements to the Daily Weather Reports in the newspapers, giving readers a glimpse of the future, such as: “General weather probable the next two days in the North - moderate westerly wind; fine.” His orders from the Board of Trade made clear that his office was not to make “predictions,” so FitzRoy coined a new term for these statements: forecasts. Unlike prophecies or predictions, he said, a forecast is “the result of scientific combination and calculation.”

From empirical evidence to models

FitzRoy’s forecasts may seem limited now, but they quickly became a popular section of the newspaper—the same way that weather forecasts are an indispensable part of the news today. However, these early forecasts lacked a theoretical or mathematical basis. Like Maury’s maps, they were based on empirical evidence: the tens of thousands of observations the office had collected over many years. To make his forecasts, FitzRoy assumed that the processes that had happened in the past were likely to happen again. In other words, he used movements of past storms to forecast how a current storm would move. This general approach often worked when looking ahead a day or two, but it was limited in scope and scale and hard for anyone less experienced to replicate.

Meteorologists would require more than empirical evidence and experience to develop longer-term, more precise, and more reliable weather forecasts. To be more reliable, forecasts needed grounding in a deeper understanding of the physical processes in the atmosphere so that computational models could be developed to move the processes forward in time. Those models also needed to be informed by data throughout the atmosphere, not just data collected from ground-based stations.

In the 1860s, the British turned their attention to balloons to gather those data, with James Glaisher again leading the way. In 1862, Glaisher took a trip to the upper atmosphere with an experienced balloonist, Henry Tracy Coxwell, as captain (Figure 4). He took regular readings from his thermometer, barometer, and hygrometer, documenting the decreasing temperature and pressure as their altitude increased. As they ascended above 27,000 feet, Glaisher passed out from lack of oxygen. Coxwell had lost control of the line that allowed him to release gas to lower the balloon, and they are thought to have reached a height of approximately 37,000 feet before he could untangle the line and begin the descent. Glaisher regained consciousness after about 10 minutes and started taking readings again. Theirs was the highest ascent at the time and the first to take detailed measurements along the way, offering a new view into the upper atmosphere.

Throughout the rest of the 1880s and 1890s, the practice of collecting data from the atmosphere expanded through the use of balloons and box kites carrying instruments. The early 1900s saw the use of unpiloted balloons that allowed for simultaneous launchings and measurements, reflecting Glaisher’s vision for his on-the-ground network. The expansion of the network and the vast amount of data collected paved the way for developing broader understandings: theories and models that could explain why the atmosphere behaved the way it did, and the ability to use those theories and models to forecast the weather.

The spark for broader understanding did not come within the meteorology community, however, but from physics. In the late 1890s, the Norwegian physicist Vilhelm Bjerknes was struggling to achieve the recognition and attention he believed his work in hydrodynamics and electromagnetics deserved. Yet, he was intrigued by the data collected with balloons and saw how his physics-based circulation models could be applied to the atmosphere and oceans. As he participated in efforts to standardize collecting and reporting methods for atmospheric data, he developed a vision for how the equations governing flow could be used in weather prediction.

In 1913, Bjerknes was hired by the University of Leipzig in Germany to lead a new geophysics institute. He moved there with most of his family from Kristiana (as Oslo, Norway, was then called). His son Jacob, who was 15, stayed in Norway to finish his schooling. A year later, in 1914, World War I began, and Germany was engaged in fighting on several fronts. The university’s German students, staff, and the rest of the country’s men were called into service. One of Vilhelm’s doctoral students, who was studying the “lines of convergence” often accompanied by damaging thunderstorms, was killed in battle. Afterward, Vilhelm brought Jacob (now 17) from Norway to help him and take over the work.

Jacob mapped the wind directions measured at observing stations across Europe over days and weeks. By doing so, he demonstrated that winds converged (flowed generally towards each other) or diverged (flowed roughly away from each other) along narrow, linear zones that were thousands of kilometers long. Because he looked at observations from the same stations over time, he could also see that these convergence and divergence zones moved over time (Figure 5). In 1917, Jacob published a paper describing “convergence zones” as places where air was rising and clouds and precipitation were likely to form, whereas “divergence zones” were places where air was sinking and the skies were clear (Bjerknes, 1917).

By the time this paper was published, the war had made living conditions in Germany increasingly difficult, and Vilhelm’s family and colleagues were concerned for their health and safety. A colleague at the Bergen Museum in Norway urged him to leave Leipzig and establish a new geophysics institute in Bergen. In the summer of 1917, the family and research team returned to Norway. Moving from an established research university to a non-academic museum during wartime required Vilhelm to rethink his vision. He would no longer have students or a research infrastructure to support the theoretical weather predictions he had been pursuing. As a result, he shifted his focus to practical weather forecasting in service to agriculture. To prepare a summer agricultural forecast, he asked the Norwegian government to increase the number of observation stations significantly. Their work began in earnest in 1918, producing three forecasts a day for three locations in Norway: Bergen, where Jacob was the forecaster, Trømso, and Kristiania (Oslo).

The additional data also helped the team pursue its theoretical work. Jacob again identified convergence zones and tracked their movement across Norway. Importantly, he was also integrating data from higher in the atmosphere. With more data, it became clear that the concepts of convergence and divergence lines were not sufficient to explain the processes in the atmosphere. Instead, he considered these as discontinuities, or abrupt changes in air temperature and/or wind direction. He connected these discontinuities to each other in larger systems called cyclones, in which the winds followed a roughly circular motion. He developed a cyclone model, shown in Figure 6, that synthesized all of his observations (Bjerknes, 1919).

Jacob depicted a cyclone consisting of a warm sector and a cold sector. The sectors were separated by a “steering line” at the front of the warm sector, marked by a wide zone of clouds and precipitation, and a “squall line” behind the warm sector, marked by a narrower zone of tall clouds and intense precipitation. The winds move counterclockwise, and the whole system moves to the east. He also includes two cross-sections through the cyclone, showing that these lines at the surface are gently sloping boundaries between warm and cold air. With this diagram, Jacob went beyond just assembling his observations: He had developed a model based on his observations that described how cyclones formed, moved, and changed over time.

To test this model, the Bergen team expanded their observations to include historical data from ships crossing the Atlantic. In the summer of 1920, an American graduate student at Berkeley, Anne Louise Beck, received a scholarship to spend a year studying and working with the Bergen team. She participated in creating the forecasts three times a day. Beck reported that the Norwegians’ forecasting method, based on their new theory about the interactions of polar and equatorial air, was entirely different from the U.S. Weather Bureau’s method. She was both witness to and participant in the ground-breaking work.

In 1921, Vilhelm published the new theory. He began, “The great changes of the weather in our latitudes have been found to depend on the passage of a line of discontinuity, which marks the frontier between masses of air of different origin” (Bjerknes, 1921). He described the two air masses as:

• Air of polar origin, which was of low temperature for the latitude, dry, and of high visibility, with prevailing motion from the east and north
• Air of tropical origin, which was generally higher temperature, humid, and hazy, with prevailing motion from the west and south

What Vilhelm described as the “line of discontinuity” marked the boundary between these two air masses. In the midst of World War I, and having left the battlefronts of Germany behind, he called this boundary the “polar front-line,” where the cold air mass “battled” the warm air. By assembling data from across the Atlantic, the team of researchers was able to show that the polar front-line extended halfway around the world and was neither stationary nor straight (Figure 7). Based on their model, Vilhelm hypothesized that it continued around the world to form a complete circuit.

Finally, here was the theoretical basis for weather forecasting, the explanation for the empirical data that had been collected on ships, on land, and in the air over a hundred years. This set of papers, led by the father-son team of Vilhelm and Jacob Bjerknes, established what became known as the “Norwegian Cyclone Model,” which explains the processes in the atmosphere that drive weather and storm generation in the temperate zone of the northern hemisphere. Key to the model are the concepts of air masses and fronts.

Air masses and pressure systems

Jacob Bjerknes introduced the concept of an air mass in his 1919 paper by describing regions of cold air and warm air and an abrupt discontinuity between them. Vilhelm refined this concept and described two specific air masses, polar and tropical, and the temperature and humidity that defined them. Today, meteorologists have added only slightly to Vilhelm’s classification while maintaining the same definition.

An air mass is a large body of air in which the temperature and humidity are essentially uniform. Large can mean thousands of kilometers wide and several kilometers thick, reaching the top of the troposphere. The temperature and humidity of an air mass are characteristic of two major components of where it was formed:

• Maritime or continental, which is the primary determinant of the humidity (maritime = humid, continental = dry) and is indicated by a lower-case letter (m or c)
• Latitude, which is the primary determinant of temperature (tropical = warm, polar = cold, arctic = very cold) and is indicated by a capital letter (T, P, or A)

For example, an air mass that develops over the northern Pacific Ocean will be a maritime polar air mass (cold, wet, labeled mP). One that forms over southern North America will be a continental tropical air mass (warm, dry, labeled cT).

An air mass's characteristic temperature and water vapor content determine its density, which influences the atmospheric pressure exerted by the air mass at Earth’s surface. Colder air is denser, producing a high-pressure system in which air descends and spreads out at the surface; warmer air is more buoyant, producing a low-pressure system in which air converges at the surface and rises. As explored in our module Water in the Atmosphere, as air rises, it cools, and water vapor will condense, often forming clouds and potentially precipitation. In contrast, descending air becomes warmer and causes evaporation from the surface below. As a result, low-pressure systems tend to be cloudier and stormier, whereas high-pressure systems are often cloud-free.

If air masses were stationary, we wouldn’t need to do any weather forecasting because conditions would remain the same. But air masses move, pushed around by the rotation of the Earth and by high-altitude winds. As a result, air masses run into each other along their boundaries, called fronts.

What happens at fronts

What Jacob called the squall line is where a cold air mass overtakes and displaces a warm air mass, known today as a cold front. Cold fronts form in low-pressure systems, extending to the south and west from the center of the system in the northern hemisphere, as shown in the model in Figure 6 and by the blue line and triangles on the portion of the weather map in Figure 8. As the front approaches, barometric pressure drops because the warm air mass is rising as the colder, denser air mass moves in. At the front, the temperature drops abruptly, and the warm air rises along a relatively steep gradient, rapidly cooling the air and causing tall clouds to form and potentially thunderstorms. As the front passes, pressure will rise again as the cold, dense air replaces the warm air. Cold fronts typically move through an area relatively quickly.

What Jacob called the steering line is where a warm air mass overtakes and displaces a colder air mass, known today as a warm front. At a warm front, the warm air mass still rises over the cold air mass, but the front moves more slowly because the rising air is not as efficient at displacing the colder air mass. As a result, the gradient of the rise is gentler (see Figure 4), providing more warning of the approaching front. High, thin clouds usually precede a warm front, and the clouds descend in altitude as the front approaches. Precipitation is typically gentler and lasts longer than at a cold front. Warm fronts are depicted on maps with red half-circles along a red line (Figure 8).

Cold fronts move faster than warm fronts, so they often “catch up” to the warm front, displacing the warm front and putting two cold air masses in contact. This is called an occluded front, and is shown in Figure 8 by a purple line with alternating half-circles and triangles.

Fronts can also stop moving, with winds blowing roughly parallel to the boundary between the air masses and neither air mass displacing the other. This is a stationary front, marked on a weather map by alternating red half-circles and blue triangles on opposite sides of a line (Figure 8). Significant precipitation can occur over several days at stationary fronts, which may produce dangerous flooding conditions at the surface.

When Anne Louise Beck brought the new model and her forecasting skills back to the United States in June of 1921, she used them to analyze US weather data for the previous month of January and show the progression of a cold front across the country. She submitted her thesis and analyses to the Monthly Weather Review for publication in 1922. However, the editor redrew and heavily altered Beck’s maps in the final publication, distorting the cold front and pushing back against the new Norwegian model and the idea of a young woman presenting it (Beck, 1922).

The limitations of the Norwegian Cyclone Model and fronts

The warm and cold fronts generated by the motion of air masses produce most of the weather in the temperate zone of the northern hemisphere, where the rotation of the Earth and the latitude produce frequent instabilities (cyclone systems) along the polar front that the Bjerknes team identified. However, the same model does not explain the dominant weather patterns at other latitudes. Weather in the tropics is dominated by Hadley cell circulation, for example, as described in our module Circulation in the Atmosphere. Hurricanes are cyclonic storm systems that do not form along fronts but are generated by the interaction of warm ocean water and the overlying atmosphere. The development of the Norwegian Cyclone Model and its use in forecasting reflects the dominance of Western European and American scientific influence at the time and the desire of those societies to reduce economic losses on land and sea. When coupled with rapid communication, use of the model substantially improved the length and accuracy of forecasts in temperate regions.

Modern weather forecasting

The use of weather radar and the development of satellite technology in the latter half of the 20th century revolutionized forecasting. Radar allowed meteorologists to observe the structure and movement of precipitation associated with fronts, and satellites provided a global view of cloud patterns and helped track the development and movement of weather systems. Today, advanced computer models use mathematical equations to simulate the atmosphere’s behavior, including the formation and movement of fronts. The combination of numerical models and observational data has significantly improved the accuracy of weather forecasting. Every meteorologist today has access to high-resolution satellite data, a network of weather stations, and models from the National Weather Service that help them accurately forecast not just the arrival of a storm but how much the temperature will change, when and what kind of precipitation will fall, and the strength of the winds—to a point. Our models still rely on empirical evidence, and as Earth’s climate warms, we are experiencing extremes that we have not seen before and can lie outside the realm of our modeling. That doesn’t mean you should not trust the forecast for the day’s weather, but it does mean that we have more to learn about our planet and its weather.