Trends in U.S. Climate during the Twentieth Century

by Thomas R. Karl, Richard W. Knight, David R. Easterling, Robert G. Quayle

Has the climate of the United States changed significantly during the century that is about to end? In what ways and by how much? Have national trends emerged that agree--or perhaps disagree--with what is expected from projections of global greenhouse warming?

Before such questions can be answered, we need to remind ourselves that "climate", as it is defined for a specific region and time, includes more than the simple average of weather conditions. Either random events or long-term persistent change, or more often combinations of them, can bring about significant swings in a variety of climate indicators from one time period to the next. Examples include a year dominated by severe drought and the next excessively wet; a series of bitterly cold winters followed by winters more mild; one scorching summer preceded by a summer pleasantly warm; years with numerous severe storms followed by years with few severe storms. The temptation at each time and place is often to attribute any of these temporal and sometimes local variations to a wider and more pervasive change in climate. The challenge to the climatologist is to separate any meaningful signals from ever-present noise, and to discern, if possible, whether there is indeed at work the sometimes slow and subtle hand of significant change. The second task, which is even harder, is to identify, unequivocally, the cause.

In this review assessment we focus on the climate history of the conterminous U.S. (excluding Alaska and Hawaii) for the century that is about to end. We first examine the observations that conventionally describe weather and climate and next, the now pressing question of whether observed changes in this country are consistent with predictions of anticipated, global greenhouse warming.

In doing so, we need always remember that while we are among the largest countries in terms of land area, the U.S. covers but about 2% of the surface of the Earth; what happens meteorologically within its borders is but a sampling of that amount drawn from an interrelated, global system. While we have a U.S. National Climatic Data Center (NCDC), its mission is necessarily global because climate itself is a global phenomenon: the atmosphere respects no national boundaries. At the same time, it is local weather conditions that most affect us, individually or collectively. Any response to the question of impending climate change must first be weighed in light of regional and national impacts, in this country or any other.

The temperature of the air is controlled to large degree by effects of naturally-occurring gases such as atmospheric water vapor. Other radiatively-active gases whose concentrations have changed due to human activities include carbon dioxide, methane, nitrous oxide, ozone, and certain compounds of carbon with fluorine and chlorine, called halocarbons. Though all are minor constituents of the atmosphere, most of them are long-lived in their effects: once there, they remain for decades to centuries. As is well-documented and now well-known, these so-called "greenhouse" gases have all been markedly increasing in amount since about the time of the industrial revolution, that began in earnest some 150 years ago. The largest and best-known contributor is carbon dioxide, originating principally from the burning of wood and coal and petroleum derivatives.

Mathematical climate models now indicate that a doubling of the present level of atmospheric carbon dioxide will, in time, raise the mean global surface temperature by 1.5 to 4.5°C (about 3 to 9°F), depending on the model used. Using a variety of assumptions about global population growth, availability of fossil fuel, and global economic growth, carbon dioxide concentrations are projected to increase between 75 and 220% by the end of the next century.

Measurements of past and current levels of carbon dioxide and other greenhouse gases indicate that we should have already increased the global greenhouse effect by man-made, or anthropogenic additions, by nearly 40% in the last 150 years. If these changes were the only process of importance, then the same mathematical climate models suggest that the average global surface temperature should have risen by about 1°C during this time. Available climate data suggest that the mean global temperature has indeed risen, but unsteadily and by only about half that amount.

In addition to the increase of greenhouse gases however, we have also changed the composition of the atmosphere in ways that act to cool the surface temperature. This includes the anthropogenic decrease of stratospheric ozone, and an increase in anthropogenic microscopic sulfate particles, often readily apparent during the warm season as smog. The effect of these additional atmospheric constituents on global climate is less certain than that of the better known greenhouse gases, but models suggest that in some areas they may have already acted to significantly retard greenhouse warming. It is important to note, however, that the global-scale warming predicted in climate modeling experiments from future greenhouse gas increases is substantially larger on a global average than the regional cooling expected from these other sources.

Efforts to identify unequivocal effects of greenhouse gas warming are best studied through global analyses. Such analyses have been made and also assessed in both intergovernmental and national reports that address the very practical question of climate change. Three of these are listed at the end of this review. All of the commissioned assessments have concluded that observed changes in global climate are not yet sufficiently large to be ascribed unequivocally to anthropogenic increases of greenhouse gases. At the same time, anticipated greenhouse warming might be expected to affect the climate of this country in several recognizable ways. These changes, in rough order of our confidence in our projections include:

• An increase of mean surface temperature, more pronounced during the cold season;
• An increase in precipitation, especially during the cold season;
• More severe and longer lasting droughts, particularly during the warm season;
• A small, but significantly greater increase of nighttime temperature compared with daytime temperature, primarily during the warm season;
• A greater portion of warm season precipitation derived from heavy convective rainfall (showers or thundershowers) compared with gentler, longer-lasting rainfalls; and
• A decrease in the day-to-day variability of temperature.

In addition, if all other factors were constant, warmer sea-surface temperatures might be expected to increase the severity and/or frequency of hurricanes affecting the United States and adjacent waters, although the natural variability of hurricanes is so great that such an effect would be very difficult to identify.

Confounding any search for anthropogenic effects are the natural changes and variations of climate that will constantly add to or subtract from the expected signal. Examples include changes in upper atmospheric steering winds (commonly known as the jet stream) due to ocean-atmosphere interactions; changes in the circulation of the ocean that can influence air temperatures; effects of major volcanic eruptions; feedbacks from changes in the land surface, as in soil moisture, snow cover, and plant cover; and changes in the energy received from the Sun.

What is certain is that our ability to distinguish significant climatic change must begin with current and past measurements that are and have been collected in this country and around the world.

The climate data that we have compiled and analyzed here are drawn from the U.S. National Climatic Data Center in Asheville, which manages all U.S. atmospheric data that have been recorded since the advent of scientific instruments. The collection is immense and constantly growing, with enough data on-hand today to fill to capacity about half a million average-size, 1995 personal computers. Many of these measurements have been collected for the specific purpose of daily weather prediction, and not to monitor long-term climate change, which requires a sometimes different set of controls. For this reason the NCDC, in cooperation with other agencies, has undertaken numerous scientific studies to remove possible artifacts in the data collected, such as the effects of changing instruments, of local landscape or urban changes, or of changes in observing procedures.

When we take spatial averages of the total annual rain and snowfall at each of thousands of stations that record precipitation each day in the U.S. we see fairly large decade-to-decade variations (Fig. 1). Although there is an absence of any continuous end-to-end trend, since about 1970 precipitation has tended to remain above the twentieth century mean, averaging about 5% higher than in the previous 70 years. Such an increase hints at a change in climate. Statistical analysis suggests that the change is unusual, but there is still about a 10% chance that such a change could arise from a stable or quasi-stationary climate without any real long-term changes. The end-of-century increase is mainly due to increases during the second half of each year, particularly during the autumn. On a regional basis (Fig. 3) we see that the increase is widespread within the U.S., and that local increases of about 20% are not uncommon. The increase is not apparent everywhere, however, as some states including California, Montana, Wyoming, North Dakota, Maine, New Hemisphere, Vermont, and parts of the Southeast have experienced decreased precipitation in the same period.

Variations in precipitation strongly affect the frequency and severity of droughts and excessive wetness. A common index used to quantify long-term moisture anomalies in the U.S. is the Palmer Drought Severity Index or PDSI. This index categorizes moisture conditions in increasing order of intensity as near-normal; mild to moderate; severe; or extreme for drought or wetness. The PDSI is affected by both long-term moisture shortages and excesses, and by variability of temperature-driven evaporation from soils and transpiration (release of water vapor) from trees and other plants. Since warmer conditions are capable of evaporating more water from the Earth's surface, both temperature and precipitation affect the drought index, but temperature effects are less a factor than direct changes in precipitation.

To characterize long-term variations of drought or wetness we can calculate the proportion of the U.S. under conditions with severe and extreme (both of which we simply characterize as "severe") drought or moisture surplus, as defined by the PDSI. Considerable decadal variability of drought and wetness is revealed by these indicators (Fig. 2). The droughts of the 1930s and 1950s stand out in the upper curve of Fig. 2 as remarkable events. During 1934, the worst year, on average nearly 50% of the country was in severe drought, dwarfing by comparison the recent drought of 1988. Since about 1970 however, more of the country has tended to remain excessively wet: over 30% of the country has experienced a severe moisture surplus for at least one year in each of the past three decades. The summertime catastrophic flooding of the Mississippi River and its tributaries during 1993 is an obvious example of these severe moisture surplus events.

The effects of a long-term moisture deficit or surplus are generally proportional to the area covered in either severe drought or severe moisture surplus. If we consider the sum of the proportion of the country in either of these severe categories, no systematic trends are evident in the present century, although during the past few decades there has been a tendency for a greater portion of the country to be either in severe drought or severe moisture excess.

An analysis of precipitation changes would be incomplete without consideration of changes in daily precipitation events. During the century the average number of days per year with precipitation has increased, in a trend that mimics the increase of annual precipitation, shown in Fig. 1, that began about 1970. Since 1970 there have been about 2% more days per year with precipitation than earlier in the century. This equates to an average increase of about 6 more precipitation days per year, but varies depending on the part of the country being examined.

We can also calculate for each year the proportion of the country that has had a much greater than normal amount of precipitation (defined as within the upper ten percent, or tenth percentile of all annual values) derived from extremely heavy 1-day precipitation events. This indicator (Fig. 4) can be reliably calculated at least back to 1910. It is clear that during the present century there has been a steady increase in the area of the U.S. that is so affected. Given the natural variability from year-to- year and thus the natural variability on the century time-scale the observed climate change could be expected, by chance, less than 1 time in 1000. The proportion of the country with a much greater than normal number of wet days has also increased much more than would be expected from random variations. Meanwhile the proportion of the U.S. with a much greater than normal number of dry days has shown little overall change. To better interpret these changes it should be recognized that for some regions and for certain months of the year, 1-day precipitation events exceeding 2 inches never occur, or there are too few wet days in a given month to establish an upper and lower tenth percentile. These areas are not considered in the indicators, but include only a small portion of the country.

Data for the U.S. describing cloud cover are reliable back to at least 1950, and indicate that cloud amount has increased over the nation by about 1% since 1970 compared with the previous two decades (Fig. 5). The prolonged drought of the early 1950s in the southern-half of the U.S. colors this picture, however, since the record begins during a time that was unusually dry and cloud deficient. Nonetheless, the increase in cloudiness is most pronounced during the autumn, and it varies in concert with the end-of-century increase of precipitation noted earlier. It is also related to a systematic lowering of cloud ceilings (the height above ground of the bottom of the cloud) that naturally accompanies the onset of precipitation. Increases of cloud amount have been observed both during the day and night. Expected reductions in sunshine have also been measured during the same times that cloudiness has increased.

A straightforward statistical average of mean temperatures across the U.S. gives evidence of a rise through the century of about 0.3 to 0.4°C (0.6 to 0.8°F), although so crude a characterization of mean temperature change in the U.S. would be indeed a gross oversimplification. The record, shown in Fig. 6, shows a sharp rise in temperature during the 1930s and a modest cooling trend from the 1950s to the 1970s, at which time the temperature rapidly increased, and has since remained as high as some of the very high temperatures that were recorded during the droughts of the 1930s. It should be noted however, that the more recent warmth is accompanied by relatively high amounts of precipitation, unlike the dry 1930s. Although U.S. temperatures have substantially increased, as has precipitation, the increase is neither large enough, nor temporally consistent enough, to completely dismiss the notion that the change may have arisen due to natural variations, which could occur through chance about 1 time in 20.

The increase in annual temperatures after the 1970s is mainly the result of significant increases of temperature during the first six months of the year (winter and spring). Temperatures during summer and autumn have changed little after dropping from conditions of the warm 1930s. Unusually high precipitation and cloud amount tend to cool the air, especially during the second half of the year. It is rare to find much above normal precipitation and cloud amount during these two seasons when temperatures are higher than normal.

On a regional basis the West contributes most to the increase of annual average nation-wide temperatures. As with drought and excessive moisture, portions of the country can be extremely cold at the same time that others are unusually warm, leading to an average national temperature that is near-normal. Similarly, abnormally high daytime maximum temperatures can occur while nighttime temperatures remain below normal, or vice-versa, although these are not usually the case.

The warmth of the 1980s and early 1990s is better reflected in the mean daily minimum temperature compared to the maximum temperature. The lower curve in Fig. 7 shows that the proportion of the U.S. with much above normal daily minimum temperatures (within the upper tenth percentile of the local, long-term record) has sustained itself at high levels since the late 1970s. In contrast, the area of the country affected by much below normal daily minimum temperatures, shown in the upper curve, has remained relatively small. Unlike the much below normal category for the minimum temperature the proportion of the country with much below normal maximum temperatures (not shown) has been fairly steady since the 1930s. The recent increase of the minimum temperature relative to the maximum temperature is directly related to an increase in cloud amount over the past few decades. However, the area of the U.S. separately affected by both much above normal maximum or minimum temperatures has increased since the 1970s.

The fraction the country with either much above or much below normal monthly average temperatures has changed little in the course of the century. The tendency for a larger area of the U.S. to have much below normal temperatures in the early part of the century has been balanced by the opposing case of much above normal temperatures in the last few decades.

Significant changes in temperature variability can also be reflected in the day-to-day changes of temperature. Examination of the magnitude of day-to-day temperature change for the present century indicates that there has been a rather steady and significant decline of day-to-day differences of temperature in the U.S. This reduction in day-to-day temperature swings is apparent throughout much of the country, especially during the warm season. Moreover, if each days highest and lowest temperature from many weather stations are categorized by season, the highest 1-day seasonal temperatures tend to decrease during the twentieth century, while the lowest 1-day seasonal temperatures tend to increase. Taken separately, either of these changes could occur in a stable climate, but when considered together they indicate a definite reduction in the absolute seasonal range of temperatures: a trend that is not characteristic of an unchanging climate.

Changes and variations of destructive storms are of particular interest because of their socio-economic and biophysical impact. Reliable records of the number and intensity of tropical hurricanes that reach the U.S. go back to at least 1900. Fig. 8, based on a commonly used classification of hurricane intensity, indicates that the frequency of these violent storms that make landfall in the U.S. has been relatively low over the past few decades, as compared to the middle of the century. The decline is reflected in both the total number of hurricanes making landfall in the U.S. and in the occurrence of more destructive storms. It is difficult to discern any long-term trend however, since the frequency of hurricanes was also low in the early part of the century. Furthermore, recent studies indicate that even if significant greenhouse induced warming were to occur, it is doubtful whether increases in tropical storms would be detectable due to the large natural variability in these storms.

Many of the variations and changes of surface climate are forced by atmospheric circulation changes, some of which are well documented over the past few decades. These variations in the directional movement of large masses of air are very much affected by the exchange of energy between the atmosphere and the ocean. The clearest examples are the El Niñ o phenomena of the equatorial Pacific Ocean. Since the winter of 1976-77, the frequency and intensity of El Niñ o events have increased relative to previous decades. During these years sea-surface temperatures in the central and eastern equatorial Pacific have remained anomalously warm. Such events have been directly linked to increased precipitation in the southeastern U.S. and warmer than normal temperatures in the Pacific Northwest. During these same years a large-scale redistribution of atmospheric mass has taken place in the North Pacific, associated with a change of the upper-level steering winds over the North Pacific and North America. El Niñ o events (and their opposition phases, La Niñ a events) have been quantitatively linked to the 1988 drought, to increased precipitation in the South, and to other abnormal temperature conditions in the U.S. Variations in the circulation of the North Atlantic Ocean have also directly influenced the eastern U.S. climate in the form of stronger than normal winds over these regions that seem to oscillate on decadal time-scales. Such oscillations have been linked to colder than normal temperatures in the region.

Most readers will by now agree that it is difficult to draw a simple picture that summarizes the many parameters and multidimensional aspects of observed climate change and variability, no matter how complete the record. One approach toward simplification might be to consider only long-term measurements of a few near-surface conditions: temperature and precipitation, for example, are two primary elements of climate that affect many aspects of our lives. But neither tells the whole story.

An index that combines a number of climate indicators can provide a convenient tool to summarize the varying states of climate. To be useful it must have a clear meaning, a moderately long history, and continuity into the future. Nor can it smooth out potentially important aspects of climate change in the name of intended simplification. Two types of indices have been developed at the NCDC. The first is aimed at assessing changes and variations of climate extremes, and is most relevant for gauging the potential impact of long-term climate variations and changes on natural and man-made systems in the U.S. The other focuses on changes that have been projected to occur in the U.S. due solely to anthropogenic increases in greenhouse gases.

The U.S. Climate Extremes Index (CEI) is the annual average of five indicators, where in each case we define much above or much below normal as falling within the highest and lowest tenth percentile of the local, long-term record. These are:

(1) The sum of:

a) Percent of the U.S. with maximum temperatures much below normal.

b) Percent of the U.S. with maximum temperatures much above normal.

(2) The sum of:

a) Percent of the U.S. with minimum temperatures much below normal.

b) Percent of the US with minimum temperature much above normal.

(3) The sum of:

a) Percent of the U.S. in severe drought (equivalent to the lower ten percentile) based on the PDSI.

b) Percent of the U.S. with severe moisture surplus (equivalent to the upper ten percentile) based on the PDSI.

(4) Twice the value of:

the percent of the U.S. with a much greater than normal proportion of precipitation derived from extreme 1-day precipitation events (more than 2 inches or 50.8 mm).

(5) The sum of:

a) Percent of the U.S. with much greater than normal number of days with precipitation.

b) Percent of the U.S. with much less than normal number of days with precipitation.

In any given year each of the five indicators has an expected value of 20% in that 10% of all observed values should fall, in the long- term average, in each tenth percentile, and there are two such sets in each indicator. An extremely high value in any one of the five indicators does not exclude extremely high values for the others. The fourth indicator, related to extreme precipitation events, has an opposite phase that cannot really be considered extreme. Namely, the fraction of the country with a much below normal percentage of annual precipitation derived from extreme 1-day precipitation amounts. For this reason the fourth indicator is multiplied by two to give an expected value of 20%, comparable to the other indicators.

Overall, the CEI gives slightly more weight to precipitation extremes than to extremes of temperature. A value of 0%, for the CEI--the lower limit--indicates that no portion of the country was subject to any of the extremes of temperature or precipitation considered in the index. In contrast, a value of 100% would result were the entire country under extreme conditions throughout the year for each of the five indicators--a most ominous but improbable scenario. The long-term variation or change of this index represents the tendency for extremes of climate to either decrease, increase, or remain the same.

The 80-year record of the CEI depicted in Fig. 9 demonstrates that the climate of the U.S. in this period has included large decadal fluctuations of climate extremes. Since about 1976, the time when the atmospheric circulation over the Pacific and North America underwent a significant change, the CEI has averaged about 1.5% higher than the average of the previous 65 years. This is equivalent to a persistent increase of extreme events covering an area somewhat larger than the state of Indiana. Other notable times of extreme climate variations include the 1930s and 1950s, but the more recent spell of extreme climate is of longer duration. This increase in extremes is related primarily to the increase in three of the five indicators: the frequency of long term drought severity and moisture excess; the frequency of extreme 1-day precipitation events; and a much greater than normal number of days with precipitation. The increase in climate extremes over the past 15 to 20 years is not, however, of sufficient persistence and magnitude to suggest that the climate really has changed. Such an increase, due simply to natural year-to- year variability or to El Niñ o events, is not unexpected.

The U.S. Greenhouse Climate Response Index (GCRI), is based on the set of anticipated greenhouse climate response indicators that were listed earlier in this review, and is intended as a means of possible early detection and monitoring of anticipated greenhouse-induced climate change as applied to conditions in the U.S. Other anthropogenic influences on climate, such as the cooling effects of sulfate-induced smog, as well as natural climate change mechanisms, will either enhance or reduce the GCRI. It is worth noting however, that in the U.S. the net change of anthropogenic emissions of sulfur dioxide (leading to sulfate-induced smog) was negligible between 1950 and 1993.

The U.S. GCRI is calculated from the arithmetic average of four indicators:

(1) The percent of the U.S. with much above normal minimum temperatures.

(2) The percent of the U.S. with much above normal precipitation during the months October through April (the cold season).

(3) The percent of the U.S. in extreme or severe drought during the months May through September (the warm season).

(4) The percent of the U.S. with a much greater than normal proportion of precipitation derived from extreme 1-day precipitation events (exceeding 2 inches or 50.8 mm).

In each case, we define much above normal conditions as those falling in the upper tenth percentile of the local, century-long record. Each of the fo