[illustrations, footnotes and references available in PDF version] 

For the Full Report in PDF Form, please click here.

Summary for Policy Makers

The observations detailed herein illustrate that climate variability from year-to-year and decade-to-decade plays a greater role in Kentucky’s climate than any long-term trends. Such short-term variability will continue dominate Kentucky’s climate into the future.  At the century timescale, Kentucky’s climate shows no statically significant trend in statewide average annual temperature, and a weak tendency towards increasing statewide total annual precipitation and decreases in the frequency and/or severity of droughts. Cleary, in Kentucky over the course of the past century or so, local and regional processes have been more important than global ones in determining local climate and local climate variations and changes. The same is true for extreme weather and weather-related events—tornadoes, wildfires, and floods show no trends that could be related to “global warming,” instead, as history shows, these types of events have marked Kentucky’s past, and will continue to occur into the future. And climate change is shown to have little, if any, detectable impacts on the overall health of Kentucky’s population. Instead, application of direct measures aimed at combating the negative impacts of heat waves and vector-borne diseases prove far and away to be the most efficient and effective methods at improving the public health. 

I. Observed Climate Change in Kentucky

A. Temperature

Averaged across the state of Kentucky, the long-term annual temperature history shows no long-term trend over the past 112 years—the time since widespread records were first complied by the U.S. National Climatic Data Center.  A string of recent warm years in Kentucky, which began in 1998 followed on the heels of an extended cool period during which 29 out of 40 years experienced colder temperatures than average. Prior to about 1955, it was not unusual for statewide annual average temperatures in Kentucky to be as warm, or warmer, than temperatures typical of the past decade. In the century-long context, recent temperatures have not been at all remarkable. In fact, the temperatures during the most recent 50 years (1957-2006) were cooler on average than those of the 50 years prior to that (1907-1956).  Obviously, “global warming” has not had much of an impact on the temperatures in Kentucky.

If you examine Kentucky’s statewide temperature history within the four seasons, you again find no evidence of any “global warming” throughout any portion of the year. There has been a slight warming tendency during the winter and spring season, but slight cooling tendencies during the summer and fall. Overall, there are no strong trends and recent temperatures are unremarkable in every way when set against long-term observations.

B. Precipitation

Averaged across the state of Kentucky for each of the past 112 years, statewide annual total precipitation exhibits a slightly increasing trend of about 5% per century (which has resulted on average in about 2-3 more inches of precipitation per year). But more characteristic than a long-term trend is the large variations in annual precipitation amounts from year to year. For instance, the driest year on record in Kentucky was 1930 when only 29.39 inches of precipitation fell, while, the wettest year, 1950, received 62.93 inches. Recent annual totals show nothing out of the ordinary when compared to the observed historical record.

C. Drought

As is evident from Kentucky’s long-term precipitation history, there are sometimes strings of dry years, for instance, the early-1950s.  Several dry years in a row can lead to widespread drought conditions. However, as is also evident from Kentucky’s precipitation history, as we discussed, there has been a slight trend towards increasing precipitation amounts across the state. Consequently, the long-term trend in drought conditions in Kentucky (as indicated by the history of the Palmer Drought Severity Index (PDSI)—a standard measure of moisture conditions that takes into account both inputs from precipitation and losses from evaporation) is also towards wetter, rather than drier, conditions. In general, however, the Kentucky PDSI history is dominated by shorter term variations which largely reflect the state’s precipitation variability. Droughts in the early-1930s and mid-1950s mark the most significant events of the past 112 years.

While concerns are often raised that global warming will have negative consequences on Kentucky’s agriculture and forestry by altering the patterns and increasing the frequency of drought, as can been seen from the historical observations over the past 112 years, any changes to Kentucky’s moisture regime have been slight and towards wetter rather than towards drier overall conditions.

A history of drought in Kentucky can be traced back even further than we have direct rainfall measurements by utilizing information stored in the annual pattern of tree rings collected from area. For instance, analyzing tree-ring patterns, Dr. Edward Cook and colleagues were able to reconstruct a summertime PDSI record for central Kentucky that extends back in time more than 1500 years.  That paleoclimate record of moisture indicates that there have been extended periods in the past that have been substantially drier than present-day conditions, for example, a dry period that latest for nearly four centuries is evident in the reconstructed drought history from about 600AD lasting through about 1000AD.  The averageKentucky during that period were drier than even the worst modern-day droughts. More recently, during the past several centuries, alternating multi-decadal periods of wet and dry conditions have occurred with regularity and demonstrate that recent conditions are anything but unusual. moisture conditions in

II. Extreme Events in Kentucky

A. Fire

Along with the cycle of wet and dry periods, follows the cycles of fires. With so much recent attention given to the wildfires of California and the western United States, it is instructive to examine the history of fires in Kentucky. According to the Kentucky Division of Forestry, wildfires burn about 81,000 acres of land a year in the state, with humans being the cause of more than 90% of all fires.  Typically, more fires burn during drier conditions.  This is evidenced by the history of the numbers of acres burned each year in Kentucky.  The worst year for wildfires was 1952, in the middle of one of the worst droughts in Kentucky during the 20th century.  Other periods of high burn totals during the past 60 years include the dry conditions in the early 1980s and late 1980s and the dry years on either side of the year 2000.  But, as is also clear from the state’s fire history during the past half-century, there is not a trend towards more or worse fires—a characteristic that fits nicely with the observations that neither has there been an increase in drought conditions during the same period.

B. Floods

In addition to droughts and wildfires, the climate and geographic location of Kentucky is also one which promotes the occurrence of flooding as its northern and western borders are defined by the Ohio and Mississippi rivers.

According to the United States Geological Survey (USGS),

Floods in Kentucky are caused either by frontal systems that occur in winter and early summer, air mass-type thunderstorms from summer to early fall, or tropical storms from the Gulf of Mexico from late summer to early fall. Floods associated with frontal systems are generally the longest lasting and the most widespread. Floods associated with summer thunderstorms are more local and can be of short duration. Rainfall associated with tropical cyclones is infrequent but can produce widespread flooding.

And while flooding is a concern today, as the recent flood disasters in 1997 and 2006 attest, it has also been an all-too-common and tragic event in shaping the state’s history.

We reproduce some photographs below of notable Kentucky flooding events that took place early in the 20th century to illustrate that climate change need not be invoked as playing a significant role in current and future floods. As surely as it flooded in the past, it will again in the future.

And, the biggest flood disaster of them all was The Great Flood of 1937. The descriptions and images of this event, called “the worst natural disaster in the history of the Ohio Valley” are taken from the publication “The Great Flood of 1937” published by the Cincinnati Museum (http://www.cincymuseum.org/explore_our_sites/cincinnati_history/ovh_1937flood.pdf).

This January [2007] marks the seventieth anniversary of the worst natural disaster in the history of the Ohio Valley. Heavy rains in early 1937 led to extensive flooding along the Ohio River and numerous tributaries. Although hundreds of communities suffered, the Great Flood of 1937 struck particularly hard in Cincinnati and Louisville.

In Cincinnati, the river stayed above flood stage from January 18 until February 5 and reached its crest of 79.99 feet on Tuesday, January 26. Schools, stores, theaters and factories closed. Authorities rationed electricity, suspended streetcar service, and shut off the water supply except for four hours daily.

In Louisville, the Ohio stayed above flood stage for twenty-three days. On January 27, the river crested at 57.1 feet, almost thirty feet above flood stage. More than 60 percent of the city was under water and about 230,000 of Louisville’s 350,000 residents had to evacuate their homes. Property damage exceeded fifty million dollars.

The flood resulted from unprecedented January rain throughout the region. January 1937 was the wettest month in Ohio since 1866 with a state average of 9.57 inches. Normal January precipitation is two to three inches. The highest Ohio rainfall was 14.88 inches in Fernbank, just west of Cincinnati, but rainfall in Louisville surpassed even that total. Louisville’s January precipitation was a record 19.17 inches.

Some people suggest that the number of extreme storm events, such as tornadoes will increase in the future as a result of global warming and point to recent increases in the occurrence of tornadoes in the United States as proof.  However, in the United States, as in Kentucky, the recent increase in tornado observations can be explained simply by non-climate factors such as the expanded use of Doppler radar by the National Weather Service, an increase in the number of observers (or “storm chasers”), and an increase in the population density.  Consequently, small tornadoes that were once missed are now being detected by radar and the larger observing network. The number of strong tornadoes—those less likely to have ever been missed—has not changed at all, indicating that it is not the occurrence of storms that has changed in recent decades, but instead, our ability to detect them.

But even though the number of tornadoes impacting Kentucky is not increasing, Kentucky still is under a threat of tornado disasters.  In fact, the largest tornado outbreak in U.S. history occurred in and around Kentucky on April 3-4, 1974. During a 16-hour period, in what became known as the 1974 Super Tornado Outbreak, 148 tornadoes touched down across 13 states, leaving 330 people dead and more than 5,000 people injured.  In Kentucky, 26 tornadoes, including some of the most violent, crossed the state, killing more than 70 people, injuring more than 1,300 others and leaving utter devastation in their paths. An F5 tornado—the most violent of storms—tracked through the town of Brandenburg, Kentucky, killing 31 people and leaving most of the town unrecognizable while several other extreme strong F4 tornadoes  struck other cities and small towns, including Louisville, Alpine, and Mt. Victory (http://www.publicaffairs.noaa.gov/storms/kentucky.html). The 1974 Super Outbreak, which occurred over 30 years ago, is evidence that natural climate disasters can and do strike Kentucky and that anthropogenic global warming need not be invoked as the scapegoat.

A. Temperature-related Mortality

On oft-repeated mantra is that “global warming” will lead to an increase in the number of people who will die during summer heat waves. However, the best science proves otherwise, demonstrating that instead of simply dying, that people are better adapting to hot conditions.

A number of studies have shown that during the several decades, the population in major U.S. cities has grown better adapted, and thus less sensitive, to the effects of excessive heat events (Davis et al., 2003ab). Each of the bars of the illustration below represents the annual number of heat-related deaths in 28 major cities across the United States.  There should be three bars for each city, representing, from left to right, the decades of the 1970s, 1980s and 1990.  For nearly all cities, the number of heat-related deaths is declining (the bars are get smaller), and in many cities in the southeastern United States, there is no bar at all in the 1990s, indicating that there were no statistically distinguishable heat-related deaths during that decade (the most recent one studied). In other words, the population of those cities has become nearly completely adapted to heat waves. This adaptation is most likely a result of improvements in medical technology, access to air-conditioned homes, cars, and offices, increased public awareness of potentially dangerous weather situations, and proactive responses of municipalities during extreme weather events.

And although there are no cities from Kentucky that were included in the Davis et al. studies, in most cities in states surrounding Kentucky that were part of the investigation, there is not a bar present at all in the 1990s (see for instance, the closest cities to Kentucky included in the studies, Cincinnati, Ohio, and St. Louis Missouri, indicated by CIN and STL, respectively, in the figure above).  This indicates that there are no statistically distinguishable heat-related deaths during that decade (the most recent one studied)—meaning that the population of those cities has become nearly completely adapted to heat waves. This is likely to be the case in Kentucky cities as well, as these cities share characteristics of the climate of those of the cities noted above. This adaptation is most likely a result of improvements in medical technology, access to air-conditioned homes, cars, and offices, increased public awareness of potentially dangerous weather situations, and proactive responses of municipalities during extreme weather events.

In general, the overall pattern of the distribution of heat-related mortality in cities across the United States indicates that in locations where extremely high temperatures are more commonplace, such as along the southern tier states, the prevalence of heat-related mortality is much lower than in the regions of the country where extremely high temperatures are somewhat rarer (e.g. the northeastern U.S.).  This provides another demonstration that populations adapt to their prevailing climate conditions.  If temperatures warm in the future and excessive heat events become more common, there is every reason to expect that adaptations will take place to lessen their impact on the general population.

In a subsequent study, Davis et al. (2004) focused not just on summertime heat/mortality relationships, but looked across all months of the year. Davis et al. (2004) found that, in Cincinnati, like in most cities across the United States, higher temperatures in the months of July and August were often associated with increased mortality, but that in the most other months, the association ran in the opposite direction, that is, cooler temperatures were associated with mortality increases. 

The figure above (taken from Davis et al., 2004) illustrates this pattern of temperature/mortality relationships for Cincinnati, Ohio, the closest city to Kentucky included in the analysis. Negative bars indicate a negative relationship between temperature and mortality, that is, the colder it is, the more people die and vice versa for warmer than average conditions (i.e., the warmer it is, the fewer people die).  Positive bars indicate positive relationships between temperature and mortality, that is more people die when it is hotter than normal, and few die when it is below normal.  For the most part, Cincinnati exhibits more negative bars than positive ones, with the cold season months showing weak negative relationships and a couple months in summer showing weak positive relationships.  Taken together, this indicates that if the future was marked by a warming climate, especially one in which winters warmed a greater degree than summers (in character, matching the pattern of warming that has been observed in the Northern Hemisphere for the past 50 years or so), that there would be a weak tendency for fewer temperature-related deaths in Cincinnati and probably throughout the state of Kentucky.

All told, however, the total annual number of direct weather-related deaths, whether as a result of cold or warm conditions, is quite small compared to the annual overall mortality.  For instance, in general, in the United States, the annual mortality rate is a bit less than 1% per year, meaning that about 8,000 to 9,000 people die each year out of every million persons. In Cincinnati, the net effect of the weather in a typical year is about 1,000 times less, or fewer than ±10 deaths per million people. So, despite proclamations to the contrary, the weather and climate have only an exceedingly small impact on overall mortality when the population at large is considered.  This is true now and will remain true into the future, no matter how the climate evolves.

B. “Tropical” Disease

Diseases such as malaria and dengue fever have been erroneously predicted to spread due to global warming. In fact, they are related less to climate than to living conditions. These diseases are best controlled by direct application of sound, known public health policies.

The two tropical diseases most commonly cited as spreading as a result of global warming, malaria and dengue fever, are not in fact “tropical” at all and thus are not as closely linked to climate as many people suggest. For example, malaria epidemics occurred as far north as Archangel, Russia, in the 1920s, and in the Netherlands. Malaria was common in most of the United States prior to the 1950s (Reiter, 1996). In fact, in the late 1800s, a period when it was demonstrably colder in the United States than it is today, malaria was endemic in most of the United States east of the Rocky Mountains—a region stretching from the Gulf Coast all the way up into Northern Minnesota—including the eastern two-thirds of Kentucky.

In 1878, about 100,000 Americans were infected with malaria; about one-quarter of them died. By 1912, malaria was already being brought under control, yet persisted in the southeastern United States well into the 1940s.  By the mid-to-late 1950s, malaria was effectively eradicated from the United States. This occurred not because of climate change, but because of technological and medical advances. Better anti-malaria drugs, air-conditioning, the use of screen doors and windows, and the elimination of urban overpopulation brought about by the development of suburbs and automobile commuting were largely responsible for the decline in malaria (Reiter, 1996; Reiter, 2001). Today, the mosquitoes that spread malaria are still widely present in the Unites States, but the transmission cycle has been disrupted and the pathogen leading to the disease is absent. Climate change is not involved.

The effect of technology is also clear from statistics on dengue fever outbreaks, another mosquito-borne disease. In 1995, a dengue pandemic hit the Caribbean and Mexico. More than 2,000 cases were reported in the Mexican border town of Reynosa. But in the town of Hidalgo, Texas, located just across the river, there were only seven reported cases of the disease (Reiter, 1996). This is just not an isolated example, for data collected over the past several decades has shown a similarly large disparity between the high number of cases of the disease in northern Mexico and the rare occurrences in the southwestern United States (Reiter, 2001). There is virtually no difference in climate between these two locations, but a world of difference in infrastructure, wealth, and technology—city layout, population density, building structure, window screens, air-conditioning and personal behavior are all factors that play a large role in the transmission rates (Reiter, 2001).

IV. References

Cook, E.R., Woodhouse, C.A., Eakin, C.M., Meko, D.M., and Stahle, D.W.. 2004. Long-Term Aridity Changes in the Western United States. Science, 306, 1015-1018.

Cook, E.R., Meko, D.M., Stahle, D.W. and Cleaveland, M.K. 1999. Drought reconstructions for the continental United States.  Journal of Climate, 12, 1145-1162.

Davis, R.E., et al., 2003a. Decadal changes in summer mortality in the U. S. cities. International Journal of Biometeorology, 47, 166-175.

Davis, R.E., et al., 2003b. Changing heat-related mortality in the United States. Environmental Health Perspectives, 111, 1712-1718.

Davis, R.E., et al., 2004. Seasonality of climate-human mortality relationships in US cities and impacts of climate change, Climate Research, 26, 61-76.

Reiter, P., 1996. Global warming and mosquito-borne disease in the USA. The Lancet, 348, 662.

Reiter, P., 2001. Climate change and mosquito-borne disease. Environmental Health Perspectives, 109, 141-161.