March 1995

Text references for this study

Executive Summary

Understanding the causes for changes in the weather day to day and year to year can be difficult because the complex systems underlying weather and climate change are not completely understood.

Recent articles in the press have reported that our weather is becoming more and more extreme and more destructive. Hurricane Andrew, devastating floods in California and the Midwest and the brutal winter storms that struck the Northeast last year are cited as the most recent signs that extreme weather events are becoming more intense and more frequent. Some people suggest that the planet is becoming warmer, largely as a result of the increased use of energy and the resulting increase in carbon dioxide and other “greenhouse” gases, and that this warming is causing weather to become more extreme.

But what are the facts? Is our weather becoming more hostile? To find out, the Global Climate Coalition asked Accu-Weather, Inc., to investigate historical weather records to determine if severe weather events are more frequent or more intense today than in the past and to uncover any scientific basis for linking “global warming” to our changing weather.

Accu-Weather examined relevant historical land, water and satellite weather data, conducted numerous personal interviews with scientists active in the field, reviewed pertinent literature on the subject and analyzed global weather data published by various organizations. Accu-Weather concludes that:

• No convincing observational evidence has been found to show that hurricanes, violent tornadoes and other extreme events are more common now than they were 50 or 100 years ago. The greater attention now paid to sever weather events may simply reflect three non-weather related facts: (1) more people live in areas that were once sparsely populated or even uninhabited; (2) local media are now able to quickly report severe weather events that are occurring, or have just occurred, in distant parts of the globe; and (3) more sophisticated weather monitoring systems and a more widely distributed population mean that extreme events in remote areas are more likely to be detected.

• The number of deaths in the United States caused by extreme weather disasters declined during the latter part of the century, but the values of property damage increased. This reflects both the improvements made in systems for detecting and providing early warning of danger, and the fact that more people are populating areas where severe weather is likely to occur.

• Average global temperatures have increased slightly within the past 100 years, but this increase falls within the limits of natural climate variability and does not necessarily signal that greenhouse gases are causing global warming. Much of the temperature increase occurred before 1940, while the majority of greenhouse emissions occurred well after 1940.

Weather Changes – The Variability of the Terrestrial Climate

Does climate change? Yes.

The Earth’s climate has changed drastically, and often over geological time. All of the reasons for these changes are not completely understood, but we know that climate change predates human history. Many theories have been advanced to explain these climate changes, including theories about changes in the sun’s energy output, the Earth’s orbit, volcanism, meteorite showers and most recently “the greenhouse effect.”

Significant long-term changes in the Earth’s climate have occurred in the past and, no doubt, will occur again. This has led some people to ask when the next significant change in climate will occur, and whether human activity has inadvertently accelerated the onset of climate change. While it is impossible to answer these questions unequivocally, studies of observational data and an understanding of theoretical issues of climate do offer some insight. Put briefly, climate changes for many reasons. While climate models project that anthropogenic greenhouse gas emissions may be responsible for recent and future global climate changes, there is no convincing observational evidence to support these projections, despite observed increases in greenhouse gas and aerosol concentrations during the last 100 years.

There continues to be considerable scientific uncertainty on a.) whether or when global warming will occur and b.) what influence such hypothesized warming would have on severe weather intensity and frequency.

Some people outside the scientific community predict apocalyptic climate changes within the next few decades. Others expect the climatological “status quo” to prevail well into the future. However, we have not found convincing evidence to support the hypothesis that extreme weather events, presumed to be associated with global warming, are already increasing.

There is no question that the Earth has been subjected to many climatological extremes over geological time, measured in thousands, even millions, of years. Numerous ice ages have come and gone, with warm and even very warm periods intertwined. Although the reasons for such changes are not completely understood, significant influences are attributed to changes in the Earth’s orbit around the sun, to changes in the energy output of the sun, volcanism and to meteorite impacts.

As noted, the climactic changes over the past million years or so have been stroking, with cycles of major glaciation and deglaciation. These temperature cycles, determined from the studies of oxygen isotopes in ice cores, show significant peaks in the glaciation intervals of one hundred thousand, forty-three thousand, twenty-four thousand, and nineteen thousand years (Imbrie & Imbrie, 1979). The generally accepted hypothesis for this quasi-periodic behavior is that orbital variations of the Earth have been the major factor that has “forced” long-term climate changes. This is referred to as the Milankovitch mechanism. These orbital variations consist of three factors: changes in the obliquity (tilt) of the Earth’s rotation axis, the precession of the equinoxes along the Earth’s elliptic orbit, and changes in eccentricity of the orbit. These variations are schematically illustrated in Extra-terrestrial variables however, are not at the heart of today’s argument about the so-called “greenhouse effect.” In Figure 2 we illustrate in graphic form the simplified model of the greenhouse effect. Simplified models such as this are often relied upon in the popular press when explaining the importance of greenhouse gas emissions. The fundamental concept is that the atmosphere is transparent to visible radiation from the sun,(A significant fraction of solar radiation is relected back out to space by clouds and the surface of the Earth) which heats the Earth’s surface. The Earth’s surface in turn heats up the atmosphere by radiating energy in the form of infrared (IR) radiation back out toward space. The IR radiation increases as the average temperature of the Earth’s surface rises. The temperature adjusts until a balance is achieved. If the atmosphere were also transparent to IR radiation, then the IR radiation produced by an average surface temperature of about -18 degrees Celsius (C) would balance the incoming solar radiation, i.e., the temperature of the Earth would be -18 degrees C based on the present rate of solar energy and the distance of the earth from the sun. However, the atmosphere is not transparent to the infrared, because of the radiative absorbing properties of such “greenhouse gases” as water vapor, carbon dioxide and methane. The energy absorbed by the greenhouse gases is partially radiated back to the Earth’s surface, increasing temperatures in the lower atmosphere (and decreasing temperatures at higher levels). This warming is called the “greenhouse effect,” for it makes the Earth’s average surface temperature +15 degrees C rather than -18 degrees C.

The main absorbers of IR in the atmosphere are water vapor and clouds. We note here that since water in its various forms is the principal absorber of IR, even if all the other greenhouse gases were to disappear, we would still be left with a significant fraction of the current greenhouse effect. Nevertheless, it is presumed that increases in carbon dioxide and other minor greenhouse gases can lead to significant increases in temperature. Atmospheric concentrations of some of the other greenhouse gases have increased in the last century. A widely held contention is that these increases will continue well into the future, as they have for the past century, thereby enhancing the greenhouse effect.

It should be emphasized that such “popular press” descriptions of the greenhouse effect are overly simplistic and have a tendency to mislead non-scientists in understanding the physics involved in climate change projections. Indeed, calculations of the incoming versus outgoing radiation energy flux show that the tropics receive more heat from the sun than they radiate back out to space. Conversely, the polar regions radiate more heat back than they receive from the sun. The simplistic model shown in Figure 2 shows the radiation process only, whereas, in fact, heat is also transported vertically and horizontally by the wind and by ocean currents to maintain thermal balance. The importance of this concept has been treated in detail by Piexoto and Oort (1992), who show that without a horizontal transport of heat from the equatorial regions toward the poles, the tropics would become excessively hot and the polar regions uninhabitably cold.

Much of the surface of the Earth, especially the oceans, also cools by evaporation. Most of the evaporative moisture ends up in convective clouds (clouds with strong upward and downward motion of the air), which carry the air and its contents upward, where moisture condenses into rain and snow. Just as evaporation takes hear away from the environment (cooling), the condensation of water vapor releases heat. The atmosphere receives heat from condensation of water vapor. The atmosphere must balance the heat deposited by convection from the surface with cooling by thermal radiation. Water vapor at the surface and condensation is a major heat transfer mechanism without which the earth’s surface would become unbearably hot (Lindzen, 1994a).

In addition to the convection and greenhouse effect, other internal climatic forcing functions can play an important role in non-geological climate change. For example, the interaction between the ocean and atmosphere is known to produce significant short-term changes in the climate over large regions of the globe. The quasi-periodic appearance of El Nio Southern Oscillation (ENSO)(ENSO events, also known as El Nino, are naturally occurring phenomena that occurred before and have continued to occur after the significant increase in greenhouse gases.) is a prime example of this, and the short-term impact of El Nio on global weather has been well documented for several decades. Similarly, isolated volcanic activity also appears to influence the global climate for relatively short periods of time. The recent eruption of Mount Pinatubo illustrates the short-term cooling effect of volcanic eruptions on global temperatures. While current climate models have equations which attempt to mimic the important physical processes involved in the greenhouse effect (convection, evaporation, ENSO, volcanic eruptions, etc.), popular press greenhouse effect descriptions generally ignore these important details.

Climate Change Over the Past Century (The Observations)

Do observed data indicate significant global temperature changes?

No. Global air temperatures as measured by land-based weather stations show an increase of about 0.45 degrees C over the past century. This may be no more than normal climatic variation. However, several biases in the data may be responsible for some of this increase. Moreover, much of the observed temperature increase during the past century occurred before the rise in greenhouse gases.

Reliable global weather observations extend back to the end of the 19th century. This information is essentially confined to observations of the temperature of the air near the surface of the Earth. Detailed observations of the distribution of water in its various forms, the wind and other elements both at the surface and higher levels of the atmosphere are available only for the past half century. Consequently, searches for signals of climate change over the past one hundred years have been restricted essentially to the observations of temperatures at specific stations around the world, the greatest density of which are on land, in populated areas such as North America and Europe. Before we present the observations involved in the climatic change to date, it is important to consider the general problem of “bias” in such data. The problem with generating “unbiased” air temperature and sea surface temperatures is described in detail in “Climatic Change – The IPCC Scientific Assessment 1992” and numerous other sources (Jones, 1990a). We summarize the sources of bias below.

Biases In Air Temperature Data

The IPCC Scientific Assessment provides an important discussion of problems associated with the surface air temperature record. It states that “a number of problems may have affected the record” (IPCC ’92, pg. 207). They list these problems as:

1. spatial coverage of the data is incomplete and varies greatly;

2. changes have occurred in observing schedules and practices;

3. changes have occurred in the exposures of thermometers;

4. stations have changed their locations;

5. changes in the environment, especially urbanization, have taken place around many stations.

Much effort is expended by the scientific community in attempting to remove such effects from the data. There are several potential sources of bias:

a.) Uniform Standards “There is no international standard for the calculation of mean daily temperature. Few countries have even maintained the same practice over the past century. These changes introduce bias into the record. Many of these biases can be significant. The IPCC states that some of these biases have been corrected in existing global records, “but some have not.” They also state that these biases can be significant. (IPCC 1992, pg. 207-208)

b.) Exposure Effects “Substantial systematic changes in the exposure of thermometers have occurred. Because thermometers can be affected by the direct rays of the sun, reflected solar radiation, extraneous heat sources and precipitation, there has been a continuous effort to improve their exposures over the last 150 years. Additional biases must accompany these changes in the thermometric record. Since many of the changes in exposure took place during the nineteenth and early twentieth centuries, that part of the record is most likely to be affected…The effects of exposure changes vary regionally (by country) and seasonally. Thus, tropical temperatures prior to the late 1920’s appear to be too high because of the placement of thermometers in cages situated in open sheds. There is also evidence that for the mid-latitudes prior to about 1880, summer temperatures may be too high and winter temperatures too low due to the use of poorly screened thermometers. These effects have not yet been accounted for in existing analyses.” (IPCC 1992, pg. 208)

c.) Observation Schedules “Artificial changes of temperature of either sign may exist in other parts of the world due to changes in observation time but have not been investigated.” (IPCC 1992, pg. 208)

d.) Correction Procedures “Changes in station environment can seriously affect temperature records…When reallocations occur in a random manner, they do not have a serious impact on hemispheric or global temperature anomalies, though they impair our ability to develop information about much smaller scale temperature anomalies. A bias on a large scale can emerge when the character of the change is not random. An example is the systematic relocation of some observing stations from inside cities in many countries to more rural airport locations that occurred several decades ago. There are several possible correction procedures that have been, or could be, applied…All depend on denser networks of stations than are usually available except in USA, Europe, the Western Soviet Union and a few other areas.” (IPCC 1992, pg. 209)

e,) Urban Bias “Of the above problems, increasing urbanization around fixed stations is the most serious source of systemic error for hemispheric land temperature time series that has so far been identified.” Researchers have found that urbanization can cause a bias of 0.1 degree C, even when urban areas have populations as low as 10,000. Some new rural airport locations may have suffered recently from increasing urbanization. (IPCC 1992, pg. 209)

f.) Agricultural Heat Islands Some “heat island” effects, especially during the growing season, may also have been introduced in rural agricultural areas during the past half-century. The increase in irrigation systems in many parts of the world has allowed the area extent of crops grown in “desert” areas to expand. Expansion is accompanied by an increase in local water vapor concentrations and evaporation. Wetting the ground also raises nighttime temperatures by increasing soil conductivity and raising the dew point, thus limiting the amount the temperature can drop.

g.) Sparse Data in the Southern Hemisphere Southern hemisphere ocean temperatures have always been poorly measured. Differences in existing historical data sets exist because of differences in assumptions about the mix of wooden versus canvas buckets used during the nineteenth century. (IPCC 1992, pages 210-211) An absence of data in the polar regions of the Southern Hemisphere prior to the middle of the twentieth century also introduces uncertainty into the air temperature record.

Biases in Sea Surface Temperature Data

Concerns about sparse data are even more applicable to sea surface temperatures. Moreover,

1. The manner in which water temperatures are sampled has changed considerably over the past century. The types of buckets used in making measurements were changed, as were the depths to which the bucket is lowered into the ocean. For data collected prior to 1942, this situation has necessitated the development of varying corrections. Despite these corrections, some differences in historical sea surface temperatures remain in different data sets. Since 1942, no corrections have been applied to sea surface temperature data. (IPCC 1992, pages 209-211)

2.) There are many uncertainties in the interpretation of early marine temperature records. Some researchers have used a mixture of weather-ship air temperature data and selected land air temperature data to calculate corrections. There is still concern that these corrections may be influenced by biases in the land data, including warm biases arising from the use of tropical open sheds earlier this century. (IPCC 1992, pg. 212)

Given all the uncertainties in the measurements, climatologists have done their best to reconstruct global air and sea surface temperatures; e.g. “Trends ’93” (Oak Ridge National Laboratory 1993). Figure 3 presents average annual global air temperature anomalies (Oak Ridge National Laboratory 1993) from 1880 to 1993. Figure 4 presents hemispheric data for the sea surface temperature anomalies. When the two time series are combined in Figure 5 (from IPCC 1993) the temperature increase over the record is approximately 0.5 degrees C. The most significant increase in air temperature prior to the 1970s occurred from about 1916 or 1917 to the mid-1940s. That, in turn, was followed by some cooling in the 1950s through the 1960s and some warming in recent decades.

Some climatologists have argued that the apparent rise in air temperature from the mid 1970s (shown in Figure 3) through the 1980s is proof that human-induced global warming has begun. Recent satellite observations show different global temperature trends than surface air temperature data. Unlike surface thermometers, satellites can: (a) measure the integrated air temperature over several thousand feet in the vertical, (b) measure temperatures at various levels in the atmosphere, and (c) provide nearly complete Earth coverage in as little as a day (Spencer and Christy, 1990). Experiments comparing the satellite temperatures to those measured by radiosondes have supported the accuracy of the former.

Figure 6, from Oak Ridge National Laboratory (1993) displays the lower tropospheric temperature trends (i.e., below about 10,000 feet) from 1979 through 1993, as obtained from satellite measurements. It is clear that over the past one-and-a-half decades, lower tropospheric temperatures obtained from satellites have not shown any significant increase. This is in contrast to the Wilson/Hansen and John/Wigley data (Figure 3), which show an increase. In out opinion, hemispheric satellite temperature data must be considered more representative of global atmospheric temperatures than surface temperature data, which have non-homogeneous spatial distributions and are representative of a very thin layer of air at the surface. In addition, if significant changes in lower level temperatures were occurring, corresponding changes should be reflected in the lower troposphere and should show up in the satellite data. Therefore, since no statistical trends are detectable in the satellite data, whatever is happening at the surface must have a minor effect on overall global atmospheric temperatures.

Is Today’s Weather More Extreme?

Are weather variations much more extreme today than they were 50 to 100 years ago?

No. Although some people have argued that hurricanes are becoming stronger and more frequent, that tornadoes have increased in number and that droughts and floods are becoming more common, recent work by scientists worldwide disputes this hypothesis. In fact, observational data show that the frequency of both hurricanes and violent tornadoes have not increased in recent decades. Sound theoretical arguments have been advanced that indicate even if global warming does occur, the frequency and aerial extent of hurricanes are not likely to increase.

Ever since the first simple climate models were used to make predictions of global warming a dozen or so years ago, there has been much concern in the media. The media promoted many erroneous concepts about climate change. For example, “Global Warming Unchecked” (1993), Bernard describes in detail how global warming will cause catastrophic weather changes in the United States from the West through the Great Plains, Midwest and into the East and the South. Others have predicted an even more significant effect of global warming on such often calamitous meteorological phenomena as tropical storms. This is typified by Friedman (1989) who, using theoretical arguments advanced by Emmanuel (1987-88), hypothesizes that since global warming means higher sea surface temperature, that in turn will:

A. Lengthen the tropical storm and hurricane season;

B. Extend the area over which tropical storms and hurricanes can form;

C. Cause tropical storms and hurricanes to be more intense; and

D. Cause more tropical storms and hurricanes to strike the United States.

Other writers in the popular press (e.g., Flavin, 1994) have expressed ideas that are more extreme. Some claim that global warming has already affected the world, not only seen in the (alleged) increase in tropical storm and hurricane intensity and frequency, but also in the number of floods, intense middle latitude storms and droughts over the recent past. They cite the many media reports of severe weather events from all corners of the world as evidence that global warming has already occurred.

The popular press often ignores important statistical principles when writing about the climate. For example, statements about the “average” rainfall or “average” temperature do not contain complete information about the behavior of that meteorological variable. The average value, or as it is sometimes called the mean,(The terms “normal” and “average” are used interchangeably here.) of any variable is composed of variations about that average, or, in statistical jargon, the variance(also expressed as the “standard deviation” which is the square root of the variance). Some averages are composed of large (or high) variances of the measured quantity while others may have small variations. Indeed, the average rainfall (or temperature) may be the same at two locations but the variations about that average can be completely different. In Figure 7, we show a hypothetical times series of the variation of a weather element at two different locations over the same time interval. Both have the same average, but the variations around that average are significantly different.

There is another fundamentally important concept concerning the application of statistical techniques in scientific research that needs to be noted. This concept states that great care should be exercised in assuming that there is a cause and effect relationship between two events, even if the two events occur simultaneously or in phase with one another. One must be very careful in using statistical relationships to establish or even postulate physical processes that may be causing correlations between two events, or using such correlations for predictive purposes. For example, every human being who has lived for more than a few days has either died or is on their way to death, and all of them, at one time or another, drank water. Therefore, using a causal “cause and effect” reasoning, it follows that drinking water results in death. In fact, we know that the opposite is true, that without water death is predictable in a short period of time. This illustrates the danger of using pure statistical relationships to infer the physical reasons for events occurring.

Friedman and others have assumed that sea surface temperature greater than or equal to 26 degrees C is the only condition necessary for the formation of tropical storms. In addition, it was assumed that both the real extent and the length of time sea surface temperature equals or exceeds 26 degrees C would also increase as global temperatures increase. This simplistic reasoning has led to the conclusion that the tropical storm season would lengthen and the area of tropical storm formation would expand. In other words, there would be more tropical storms over longer periods of time and over larger areas. Assumptions have also been made(The assumption was based on theoretical arguments involving thermodynamic energy transfer processes.) that as tropical storms and hurricanes mover over higher sea surface temperatures, they would increase in intensity, ergo, there would be more intense tropical storms in the future as global warming takes place.

Recently, Lighthill and Holland, et al. (1994), have shown that these simplistic arguments and assumptions are not true. Specifically, the referenced article summarizes studies of tropical storms and hurricanes conducted by the nine authors over the last decade. Based on observational studies of tropical storms in the Northern Hemisphere, the authors concluded that the intensity of tropical storms is not directly related to sea surface temperature. In fact, they found no correlation between the intensification (or maximum intensity) of tropical storms and sea surface temperature. Moreover, the Lighthill and Holland study also shows there is no correlation between the Northern Hemispheric air temperature anomalies over the past several dozens of years and the maximum intensities reached by tropical storms. (See Figure 8.)

Rather, as Lighthill and Holland et al. point out, six conditions are necessary before a tropical storm forms and intensifies. A minimum sea surface temperature of 26 degrees C is only one of those conditions. The others are:

(1) The distance from the equator needs to be at least five degrees of latitude to bring into play what meteorologists call the Coriolis Effect produced by the Earth’s rotation. This effect generates cyclonic spiraling, or the counter-clockwise rotation of storm systems in the Northern Hemisphere and clockwise rotation in the Southern.

(2) The gradient of temperature decrease with height must be large enough so that the air that has become saturated with water vapor near the so-called “eye-wall” of the storm will be able to continue to rise as it moves up into the atmosphere.

(3) Low values of the vertical wind shear(The change in direction or speed, or both, of the horizontal wind with height near the center of the storm) are needed to avoid excessive departure from a vertically symmetric vortical structure, considered necessary to maintain or allow tropical storm evolution.

(4) Relative humidity has to be high enough in the middle troposphere to avoid drying effects of the air that becomes entrained into the eye-wall of the storm system.

(5) Finally, there must be the prior existence at low altitudes of a rather substantial amount of cyclonic vorticity, which in more common language, means there needs to be a pre-existing tendency for a counterclockwise spinning component(In the Norhtern Hemisphere) of the atmosphere to be present close to the surface of the ocean.

These conditions, derived from extensive observational records, emphasize that much more than just a sufficiently high sea surface temperature is needed for tropical storm formation and intensification. Condition 3, as listed above, (low vertical wind shear) is particularly important, and this is independent of sea surface temperature.

Consition 1 imposes a lower limit on those latitudes north and south of the equator where tropical storms can form. This upper limit, between 15 and 20 degrees from the equator, is set by the lower latitudinal boundary of a region in the atmosphere where downward motion of the air generates the “trade inversion.” This region is part of a large scale circulation that meteorologists call a Hadley Cell. Thius downward motion tends to dampen any upward motion induced by heating or horizontal forcing and is therefore unfavorable for tropical storm development.

Hadley Cells are determined by the orbit and shape of the Earth and the adage that “what goes up must come down.” (See Figure 9). The Hadley Cell’s ascending branch, air rising from hot regions near the equator, is balanced by subsidence of the air in its descending branch, stretching poleward an average of 15 to 20 degrees of latitude. This branch of the Hadrey Cell is associated with a slow, downward motion of the atmosphere. Accordingly, unlike the fast ascent of moist air in the tropical storm’s eye-wall, the subsiding air in this region of the Hadley Cell loses, by radiation, much of the heat if gains by compression. Thus the gradient of temperature drop with height becomes far too low for condition 2 to be satisfied in this region.

This and other evidence presented by Lighthill and Holland et al. casts serious doubts on the validity of the hypothesis that any future increases of sea surface temperature must widen the band of latitudes where tropical storms can intensify. Because subsidence, that is the downward motion in the descending branch of the Haldey Cell, prevents condition 2 and condition 4 from being reached there is little likelihood that the area of tropical storm formation will widen significantly. In short, all six conditions for tropical storm formation must be satisfied in order for there to be more frequent or more intense tropical storms. In addition, global warming would raise the temperature of the sea surface above 26 degrees C mainly in the regions where conditions 2 and 4 cannot be satisfied. Consequently, even if global warming were to occur, it appears that there would be very little, if any, effect on tropical storm development.

In summary, Lighthill and Holland et al. present credible reasons to reject the hypothesis that more intense or more frequent tropical storms would occur, even if surface sea temperatures significantly increased in response to global warming.

As for the frequency of other extreme weather events, the evidence we have examined does not support the extremists’ claim that, recently, the number of unusual weather events has increased. Once again we refer back to our discussion of averages versus variances and note that “climate” is defined as the average, over time, of the variations of meteorological elements that we call “weather.” In most instances, the variance of meteorological elements such as temperature and precipitation in the temperate and polar regions is usually high. Consequently, extreme departures from “average” weather patterns are to be expected in these areas.

However, many people today believe the weather is more extreme than it used to be. In large part this may be due to the media’s high-speed electronic ability to rapidly report worldwide events, including the weather. For example, we know almost instantaneously when a tropical cyclone devastates Bangladesh, or that a drought is occurring in Australia. In the past, few people in the United States were aware of these events. The same phenomenon is of course occurring worldwide with the recent explosion of cable and satellite television.

The observational evidence, however, suggests that weather extremes are not significantly different today than they were in the past. Hurricane(Tropical storms that reached hurricane force, sustained winds greater than or equal to 74 mph) data are a prime example. Figure 10 presents data on the frequency of typhoons and hurricanes in the northeast and northwest Pacific Ocean during the past few decades (Lighthill and Holland et al. 1994). The data show that typhoons and hurricanes have not increased in number or intensity during the past few decades. Figure 11 (National Weather Service, 1992) summarizes the number of tropical storms that became hurricanes in the Atlantic Ocean from 1880 through 1992. It is apparent from these data that the number of North Atlantic hurricanes has not increased significantly in recent years.

Moreover, when one considers that satellite observations have been very effective in locating and tracking storms for the past couple of decades, it is possible that the number of hurricanes in the Northern Hemisphere actually may have decreased in recent years. Some storms that we detect now probably would have gone unnoticed 50 to 100 years ago.

Questions have also been raised as to the possibility that severe, localized storms, such as tornadoes, are on the increase. Figure 12 presents data by Ostby (1993) on the frequency of tornadoes observed in the 48 contiguous states from 1953 to 1993. The top curve is the sum of weak, strong and violent tornadoes. Weak tornadoes are those with peak winds less than 112 mph, while strong tornadoes have peak winds from 113 to 206 mph. Violent tornadoes contain winds from 207-318 mph. As the reader can easily see, the increase in the total number of tornadoes is due to an increase in the number of weak, but not strong or violent, tornadoes over the past 40 years. Ostby attributes the increase in reports of weak tornadoes to several factors including: greater population and public awareness of tornadoes in tornado-prone areas, storm-chasing, and the advent of the video camera. Expanding population and public awareness are especially relevant because tornadoes are small storms and in days of more thinly dispersed population, many weak tornadoes undoubtedly went unreported. As the data show there is no evidence of an increase in strong or violent tornadoes over the 40-year sample in the 48 contiguous states. In fact, there appears to be an overall downward trend of such storms during the past 20 years.

Although we have not found any evidence of a significant increase in the frequency of tropical storms, hurricanes or tornadoes in recent years, a recent study (Karl and Baker, 1995), has shown that the average temperature and precipitation have increased over the past century in the contiguous United States. Once again, the question can be raised that if the average increases, does this mean that extremes will also increase? The answer is not simple, and it refers back to our previous discussion about the variation of a meteorological variable about its average. As shown in Figure 7,, knowledge of the average does not necessarily give us information about the variations from the average. For example, the recent heavy rainfall in January 1995 in California might be used as an example of an extreme event accompanying increased average rainfall. See Table I and Figure 13 for rainfall records in Los Angeles (Civic Center) from 1878 through 1993. Table I shows the monthly record rainfall in Los Angeles over the past 115 years. As one can see, five monthly record-high rainfall amounts occurred before this century. The heaviest annual total rainfall occurred in 1884. Figure 13 presents the number of months, over 10-year intervals, with total monthly rainfall greater than 7.00 inches. The greatest number of months with rainfall greater than 7.00 inches occurred from1930-1939. The total for the five previous decades is 20, while the total number of months with rainfall greater than 7.00 inches over the most recent five decades is 19.

Thus there is no evidence from the Los Angeles records that rainfall extremes have occurred more frequently in recent times. We note that the Los Angeles data are not representative of the entire country, but the data do not support the argument that an increase in average rainfall equates to an increase in extreme events.

To further substantiate the thesis that a large departure from the average of a meteorological element does not necessarily mean the departure is “extreme” we present Table II where the Monthly and Annual Standard Deviations of Rainfall as a percentage of the average for various sections of the 48 contiguous United States is displayed for the period from 1900 to 1993. The statistics were computed by Bronson Gardner, (personal communication, 1995) from the U.S. Historical Climatological Network data (Karl et al., 1994).

As an example of how the table can be interpreted, in the Coastal Section of the Northeast (blocked) in March, 66% of the time the monthly rainfall will be within 34.3% of the average. Whereas in the South Pacific Coastal area (blocked) of the USA in June, 66% of the time the rainfall will be within 192.2% of the average. In the latter example, the average rainfall is slight, well under one inch, thus even a small amount of rainfall could be considered an extreme. As for the former case, the average is much higher, being on the order of 4 inches, consequently a very large amount of rainfall is necessary before it can be considered an extreme.

As previously noted, Karl and Baker, 1995 also show that there has been an increase in the average temperature in the contiguous United States. We have examined a sample of record-high temperatures during the past century for a geographical cross-section of the United States east of the Rocky Mountains. These locations were selected because their data would not be greatly biased by the urban heat island effect and because they represented different climate zones. In Figures 14a through 14c we present the number of daily high temperature records set in 5 year periods at Des Moines, Iowa, Augusta, Georgia and State College, Pennsylvania. Note that if a record-high temperature for a specific calendar date matches one on the same date of any previous year it is assigned to the most recent date. Thus, there is a built-in bias, albeit probably small, toward the appearance of more record highs in recent years. Nevertheless, one can see that many daily high temperature records were set at the four locations prior to the last decade or two. There is no consistent evidence in Figures 14a through 14c that record-breaking high temperatures have been restricted to or are occurring more frequently in recent years, at the locations studied.

Changes In Greenhouse Gas Emissions

Greenhouse gases in the atmosphere have increased. As a result, has significant global warming been detected?

No. The amounts of carbon dioxide, methane and other greenhouse gases have increased in the atmosphere in the last century. A rapid increase began in the 1950s. But the observed global temperature rise from 1916 to the 1940s is not in phase with this increase.

The major greenhouse gas in the terrestrial atmosphere is water vapor. Minor greenhouse gases that are produced either naturally or by human activity include carbon dioxide, methane, chlorofluorocarbons and nitrous oxide. Of these minor gases, the most important, by far, is carbon dioxide. However, even carbon dioxide absorbs only about 5 percent of the total amount of infrared radiation from the Earth, compared to 90 to 95 percent absorbed by water in the form of clouds and/or water vapor.

There is no question that the atmospheric concentrations of carbon dioxide and the other minor greenhouse gases have increased. Figure 15, presents the history of carbon dioxide, methane and nitrous oxide concentrations from 1750 to the beginning of this decade. Observations prior to 1958 were obtained from ice cores, but since then they are based on direct atmospheric sampling.

How have these trends toward increased emissions and atmospheric concentration of greenhouse gases been interpreted? Simplistic models of the greenhouse effect (see Figure 2) “predict” that rapid increases in carbon dioxide in the atmosphere during the past century should have significantly increased global temperatures, especially in recent years. Yet there is no consistent, obvious signal announcing the presence of substantial global warming in any of the data that the authors examined. There is a general consensus in the scientific community that there has been a gradual increase of about 0.45 degrees Celsius, plus or minus 0.15 degrees Celsius, in the average global temperature since the late 1800s. However, that increase is within the realm of natural variability. In fact, the 1992 IPCC Supplemental Report states that the “global mean surface air temperature has increased by 0.3 to 0.6 degrees Celsius over the last 100 years…is…of the same magnitude as natural climate variability.” It is important to note that a significant fraction of the observed air temperature increase in the last hundred years occurred between 1916 and the mid-1940s, before the rapid increase in carbon dioxide emissions. Since much of the observed global warming occurred before the steep rise in greenhouse has concentrations, the warming must have been caused by other factors.

Summary and Conclusions

What does the future hold?

No one knows for certain. However, one can expect hurricanes, tornadoes, floods and droughts to be similar in intensity and frequency to those that have occurred in the past. Also, because world population is growing and people are building in previously uninhabited areas, world governments, weather sensitive industries and others need to prepare to handle extreme weather events. Catastrophic extreme weather events will continue regardless of whether or not climate change occurs. Additionally, research is needed to improve climate models as well as our understanding of the factors that influence climate. Observational studies also are needed in order to better document the Earth’s variable climate.

Even if the Earth has warmed slightly during recent decades a modest increase in the Earth’s temperature has not caused more tropical storms, hurricanes or tornadoes. And although Hurricanes Andrew and Hugo caused extensive property damage, these storms were no stronger than past storms. In fact, the hurricane that struck the Florida Keys in 1935 was more intense than Andrew or Hugo. However, it struck a sparsely populated area and so caused considerably less property damage and claimed fewer lives than either Andrew or Hugo.

Science and technology cannot prevent all loss of life and property from severe storms, but society can track its vulnerability, forecast extreme events, communicate warnings, improve land use and design safer structures. A number of organizations are doing just that including the National Hurricane Center in Coral Gables, Florida; the Natural Hazards Research and Applications Center at the University of Colorado; The World Meteorological Organization (WMO) in Geneva, Switzerland, and the United Nations Technical Committee in New York City. Recent articles by G.F. White (1994) and G.O.P. Obasi (1994) in the Bulletin of the American Meteorological Society also present important information on research efforts aimed at reducing natural disasters.

Interestingly, the number of United States deaths caused by natural weather disasters has declined during the latter part of this century while the dollar value of property damaged has increased dramatically. The reason this is so, seems to be fairly obvious. Weather forecasters are constantly improving their ability to alert people about impending severe storms, such as tropical storms and hurricanes. Moreover, evacuation procedures have been developed to move people from threatened areas. These changes have resulted in fewer injuries and deaths.

However, people do continue to live in and move to areas prone to floods and tropical storms. Because many people want to live or vacation at the shore, new construction occurs along many vulnerable coastal areas. When a powerful storm eventually strikes these densely populated areas, catastrophic damage is going to occur, regardless of whether global warming has occurred or not.

In summary, to date, despite increased concentrations of carbon dioxide and other greenhouse gases in the atmosphere, the Earth’s climate has not warmed significantly. Moreover, we have found no convincing evidence that the number and intensity of extreme weather events has increased in recent years. Indeed, were global warming to take place, it is unlikely that potentially dangerous storms, such as hurricanes would increase in number or intensity.

In conclusion, it would be prudent for the scientific community to carefully monitor atmospheric temperatures, especially by using satellite measurement systems. Developmental work on mathematical climate models should continue. Scientists should also under take more detailed observational studies of the temporal and spatial distribution of historically anomalous weather events. Observational programs for other real and potential global warming forcing functions such as sea surface temperature, ocean circulations, and aerosols should also be implemented on a much broader scale and in more detail than they are at present.

Text references for this study

Norman J. Macdonald, M.S., is a Certified Consulting Meterologist and former Senior Meterologist at Accu-Weather. Joseph P. Sobel, Ph.D., is a Senior Vice-President at Accu-Weather. Accu-Weather is the world’s leading commercial weather firm.