THE AMERICAN ACADEMY OF MICROBIOLOGY CONVENED A COLLOQUIUM WITH A DIVERSE GROUP OF SCIENTISTS TO DISCUSS HOW CLIMATE INFLUENCES HEALTH AND INFECTIOUS DISEASES. THE COLLOQUIUM HELD ON OCTOBER 15-17, 1999, IN TUCSON, ARIZONA, WAS A FOLLOW-UP TO A 1997 COLLOQUIUM, “CLIMATE, INFECTIOUS DISEASE AND HEALTH: AN INTERDISCIPLINARY PERSPECTIVE.” THE REPORT OF THE 1997 COLLOQUIUM SERVED AS ONE OF THE FIRST REVIEWS TO HIGHLIGHT THIS EMERGING FIELD OF STUDY. IN ADDITION, KEY AREAS OF RESEARCH, COLLABORATION, and data needs were identified. The second colloquium, upon which this report is based, convened to assess the current “state of the field”—or where we have come and where we are going. In the last three years, scientists in the fields of climatology, meteorology, microbiology, medicine, ecology, epidemiology, oceanography, and space science have collaborated to study how natural climate variability affects occurrence and prevalence of pathogenic microorganisms, vectors, and disease outcomes.

The unusually strong El Niño event of 1997-1998, followed by two years of strong La Niña conditions, the worst hurricane season on record, and further debate over anthropogenically induced climate change, have all been heavily covered by the popular press worldwide. Such widespread coverage demonstrates that the potential human-scale impacts of climate variability and climate change are of interest to the public. Partly in response to growing public concern, the United States Congress, in 1997, commissioned the United States Global Change Research Group (USGCRP) to develop a National Assessment of the Effects of Climate Change. However, the connection between oceans, climate and public health has been a subject of scientific discussion for the past decade. The science has progressed largely through joint efforts put forward by investigators from different disciplines.

The ability to understand and model the complex links between environmental, climatological, and biological systems present great opportunities for prediction and prevention of disease, rather than the current approach that heavily depends on clinical cases to appear before action is taken. The ability to predict potential outbreaks, based on susceptible populations and climate variability, is especially important for certain regions in developing nations where limited resources may be more efficiently utilized in advance to prevent a health crisis. Our goal in understanding and incorporating climate signals, therefore, is to use proactive, rather than reactive, approaches to protect public health.

Scientists are continuing to connect ocean sciences with human health. By defining and understanding such interactions there are hopes to develop models that will aid in the prevention of emerging and re-emerging diseases. To advance our understanding of the relationships between climate variability and infectious diseases, the following recommendations are made. Long term, historical disease and pathogen surveillance data must be located, salvaged, and electronically archived. Continuing emphasis should be placed on prospective long-term surveillance that includes medical, ecological, and climatological parameters at periods greater than five years. Multidisciplinary training in “bioclimatology” should be encouraged at all levels—undergraduate through postdoctoral. Increased communication between disciplines is imperative, but will require a common “language,” as well as support for interdisciplinary meetings and journals that focus on climate issues in biological and medical applications. Scientific societies may be best able to accomplish this goal. Quantitative analyses should move from the detection of associations to the development of predictive models. Finally, integrated frameworks and risk assessments will be useful tools to determine susceptibility of particular regions to certain diseases under a given set of climate conditions.

Climate variability affects every region of the world. Diseases are not necessarily limited to specific regions, and a changing climate may allow pathogens or vectors to become endemic in novel regions. Therefore, although the impact of climate on health may be local in an immediate sense, it is nevertheless a global issue.

Front Matter

Funding was provided by:

• American Society for Microbiology
• Environmental Protection Agency
• U.S. Department of Energy
• National Oceanic and Atmospheric Administration
• National Science Foundation

The authors wish to acknowledge the contributions of the following individuals to this report:

• Phillip Arkin, Ph.D.
• Mercedes Pascual, Ph.D.
• William K. Reisen, Ph.D.
• Robert E. Shope, M.D.

Executive Summary

THE AMERICAN ACADEMY OF MICROBIOLOGY CONVENED A COLLOQUIUM WITH A DIVERSE GROUP OF SCIENTISTS TO DISCUSS HOW CLIMATE INFLUENCES HEALTH AND INFECTIOUS DISEASES. THE COLLOQUIUM HELD ON OCTOBER 15-17, 1999, IN TUCSON, ARIZONA, WAS A FOLLOW-UP TO A 1997 COLLOQUIUM, “CLIMATE, INFECTIOUS DISEASE AND HEALTH: AN INTERDISCIPLINARY PERSPECTIVE.” THE REPORT OF THE 1997 COLLOQUIUM SERVED AS ONE OF THE FIRST REVIEWS TO HIGHLIGHT THIS EMERGING FIELD OF STUDY. IN ADDITION, KEY AREAS OF RESEARCH, COLLABORATION, and data needs were identified. The second colloquium, upon which this report is based, convened to assess the current “state of the field”—or where we have come and where we are going. In the last three years, scientists in the fields of climatology, meteorology, microbiology, medicine, ecology, epidemiology, oceanography, and space science have collaborated to study how natural climate variability affects occurrence and prevalence of pathogenic microorganisms, vectors, and disease outcomes.

The unusually strong El Niño event of 1997-1998, followed by two years of strong La Niña conditions, the worst hurricane season on record, and further debate over anthropogenically induced climate change, have all been heavily covered by the popular press worldwide. Such widespread coverage demonstrates that the potential human-scale impacts of climate variability and climate change are of interest to the public. Partly in response to growing public concern, the United States Congress, in 1997, commissioned the United States Global Change Research Group (USGCRP) to develop a National Assessment of the Effects of Climate Change. However, the connection between oceans, climate and public health has been a subject of scientific discussion for the past decade. The science has progressed largely through joint efforts put forward by investigators from different disciplines.

The ability to understand and model the complex links between environmental, climatological, and biological systems present great opportunities for prediction and prevention of disease, rather than the current approach that heavily depends on clinical cases to appear before action is taken. The ability to predict potential outbreaks, based on susceptible populations and climate variability, is especially important for certain regions in developing nations where limited resources may be more efficiently utilized in advance to prevent a health crisis. Our goal in understanding and incorporating climate signals, therefore, is to use proactive, rather than reactive, approaches to protect public health.

Scientists are continuing to connect ocean sciences with human health. By defining and understanding such interactions there are hopes to develop models that will aid in the prevention of emerging and re-emerging diseases. To advance our understanding of the relationships between climate variability and infectious diseases, the following recommendations are made. Long term, historical disease and pathogen surveillance data must be located, salvaged, and electronically archived. Continuing emphasis should be placed on prospective long-term surveillance that includes medical, ecological, and climatological parameters at periods greater than five years. Multidisciplinary training in “bioclimatology” should be encouraged at all levels—undergraduate through postdoctoral. Increased communication between disciplines is imperative, but will require a common “language,” as well as support for interdisciplinary meetings and journals that focus on climate issues in biological and medical applications. Scientific societies may be best able to accomplish this goal. Quantitative analyses should move from the detection of associations to the development of predictive models. Finally, integrated frameworks and risk assessments will be useful tools to determine susceptibility of particular regions to certain diseases under a given set of climate conditions.

Climate variability affects every region of the world. Diseases are not necessarily limited to specific regions, and a changing climate may allow pathogens or vectors to become endemic in novel regions. Therefore, although the impact of climate on health may be local in an immediate sense, it is nevertheless a global issue.

Introduction

IT IS “FLU” SEASON, AND THE RECOGNITION THAT CERTAIN DISEASES ARE RELATED IN SOME WAY TO THE “SEASON” OR “WEATHER” IS ACCEPTED BY THE GENERAL PUBLIC. HOWEVER, THE CAUSES AND FACTORS LEADING TO THESE ILLNESSES AND THE ROLE OF THE ENVIRONMENT AND CLIMATE HAVE NOT BEEN SCIENTIFICALLY ELUCIDATED. THE EMERGENCE OF NEW INFECTIOUS DISEASES, THE REEMERGENCE OF OLD DISEASES, CHANGES IN GEOGRAPHIC DISTRIBUTION OF DISEASES AND POTENTIAL CHANGES IN CLIMATE HAVE LEAD TO A GROWING INTEREST IN FORECASTING, OR predicting, and consequently preventing disease. An understanding of the modes of disease variability as they relate to climate fluctuations is absolutely necessary to make progress towards control and improved public health protection in the future. This includes the study of both seasonal trends of disease, which have been documented in folk medicine for many years, and other cyclic or long-term temporal trends (e.g., interannual variability, El Niño-Southern Oscillation), which have been more recently recognized and for which there are data available for key pathogenic agents. Scientifically sound models to predict or forecast the “behavior” of climate sensitive diseases can be developed using interdisciplinary approaches to understand the complex dynamics between the pathogenic agent, ecological system, climate, weather and human populations.

The world is changing, and with it disease dynamics may also change due to more susceptible populations, cancer, AIDS, an aging population, poverty, war, and the threat of biological terrorism. The re-appearance of cholera in the Americas, and the appearance of West Nile Virus in the U.S. and Dengue Virus at the Mexico-U.S. border are examples of climate sensitive diseases and pathogens that have gained the attention and concern of the public and those interested in public health. Will global climate change or the changes in pathogen distribution affect the distribution and risk of disease worldwide? This question can only be answered when we better understand how the dynamics of these diseases respond to variations in global and regional climate. In the development of multi-scale models, the future holds the promise for identifying sensitive populations and regions, and determining potential changes in geographical ranges of individual pathogens or diseases with climate variability. The ultimate goal is to give susceptible regions lead-time to prepare for potential disease outbreaks, given certain climatological and sociological conditions. Obviously, climate is not the only factor contributing to the risk of certain infectious diseases. Demographic, behavioral, or socioeconomic factors may override any climatological relationship and these can be used to for appropriate adaptation. In some situations, however, climate may be a very good proxy or indicator of susceptibility and will allow preventative measures to be put in place.

Then and Now: Where did the field stand in 1997 and where does it stand now?

In 1997, a colloquium was convened by the American Academy of Microbiology in which a diverse group of scientists were brought together to discuss how climate influences health and infectious diseases. The major approach of this unique colloquium focused on what was known about the connections between weather-related variability and infectious diseases. Known climate-sensitive diseases were categorized into vector-borne, waterborne, foodborne, and airborne groups. Additionally, plant and fish diseases also were recognized. Most often, variability in the occurrence of these diseases or pathogens could be related to specific weather phenomena, namely temperature and precipitation, and thus indirectly related to climate.

One of the recommendations of the first colloquium was to use the strong El Niño event of 1997-1998 which had been predicted and was just developing as a stepping stone for studying the effects of climate variability on infectious disease and health. The National Oceano-graphic and Atmospheric Administration (NOAA), Office of Global Programs, helped to spearhead the effort to help microbiologists and epidemiologists add a climate component to on-going research. The result was the ENSO Experiment. Under this umbrella, interdisciplinary teams were able to make ENSO connections to malaria, dengue fever, encephalitis, diarrheal disease, cholera, and fecal contamination of coastal water quality, among others.

The 1997-1998 El Niño event offered a “natural laboratory” to examine the effects of climate variability as it related to ongoing research projects on disease. As global weather anomalies associated with both El Niño and La Niña events are now well documented and can be forecast well in advance, the timing of this latest El Niño allowed researchers working on many diseases in many different parts of the world to focus on one manifestation of climate variability. Better assessment of the risk of Hantavirus in the southwest of the U.S. and cholera in South America associated with climate change was forth coming.

In a 1996 meeting entitled “Global Perspectives on the Health and Ocean Resources,” held in Massachusetts, Admiral James Watkins, president of the Consortium for Oceanographic Research and Education (CORE) presented a “road map” to the untapped wealth afforded by the oceans for human health, citing their role in global climate and its predictability. Now, in addition to predictability, the interactions between oceans, climate and health have gained significant attention.

On another front, the USGCRP (U.S. Global Climate Research Group), established by the United States Congress in 1990, was charged with conducting a national assessment of effects of climate variability and change. The U.S. National Assessment of the Potential Impacts of Climate Change and Variability was commissioned by the USGCRP in 1997 to study the effects of a changing climate on 5 key sectors and 20 separate regions of the United States and its territories. The sectors include health, agriculture, forestry, coastal areas and water resources. Each sector has submitted individual reports that discuss the vulnerability and adaptability of the nation to global climate change based primarily on extensive literature review.

The health sector summarized five categories of health outcomes that are most likely to be affected by climate change which include air-related health effects, water and foodborne disease, and vector and rodent-borne disease. A summary of climate-related health effects is presented in Table 1. The group found that the data gaps and the levels of uncertainty made it impossible to formulate a definitive statement regarding health outcomes.

TABLE 1.

LIST OF KEY EFFECTS OF GLOBAL CLIMATE CHANGE AS PRESENTED IN THE NATIONAL ASSESSMENT.

The general public is concerned about an apparently changing climate and its potential affect on their health and day-to-day lives. Yet, the earth's climate is too complex to be reproduced in the laboratory, therefore traditional scientific experimentation will not necessarily provide the answers we need and new interdisciplinary approaches must be developed. It is apparent that the maturation of a new field of study, which attempts to merge the physical and biological sciences, will need guidance and resources. Thus the goal of this 1999 colloquium was to bring together participants from the 1997 colloquium and other scientists who were now actively engaged in climate-health research. Whereas the first colloquium was able to list some suspected climate sensitive diseases, the second colloquium was able to add to this list with a greater level of confidence regarding the linkages. The greater focus, however, was on data, tools, and methods that are available and would be needed to pursue the study of climate and disease risk in a more sophisticated manner for the future.

Progress in Understanding the Role of Climate and Disease

UNDERSTANDING CLIMATE VARIABILITY AND ADVANCES IN CLIMATE FORECASTING AND MODELING CLIMATOLOGISTS HAVE REPORTED ON THE EXISTENCE AND IMPORTANCE OF SEVERAL MODES OF NATURAL CLIMATE VARIABILITY AND HOW THEY INTERACT WITH ONE ANOTHER TO PRODUCE REGIONAL TO GLOBAL SCALE CHANGES IN WEATHER AND CLIMATE. AS THE INTEREST GROWS IN CLIMATE-HEALTH RESEARCH IT IS IMPORTANT THAT THE HEALTH COMMUNITY ALSO UNDERSTAND THE CLIMATE SIGNALS AND HOW WELL THEY CAN BE INCORPORATED INTO HEALTH RELATED PREDICTIONS.

During the past two years, forecasts of global climate variability have progressed from potential to realized. Regular forecasts of indices of the El Niño/Southern Oscillation (ENSO) based on both statistical and dynamical models are available. Forecasts of the probability of seasonal anomalies in temperature and precipitation are being made for all areas of the globe other than Antarctica. These forecasts are based on the products of a number of tools, including global numerical models of the ocean and atmosphere, statistical models relating regional anomalies to specific sources of variability, such as ENSO, and an understanding of the errors and shortcomings of these tools. A study comparing the forecasts issued by the International Research Institute for Climate Prediction (IRI) with observed conditions has shown that, for most continents, the forecasts exhibit greater skill than a variety of simple control processes, such as assuming that current conditions will continue.

Those of us in the health, disease and ecology fields do not yet fully appreciate the vast number of modes of climate variability and have focused on ENSO because of its notoriety, importance and predictability. In fact, nearly all of the skill in current climate forecasts originates with variability associated with ENSO. Yet in some regions other climate signals may be more important than ENSO. Unfortunately the predictability and impacts of these are less well understood. Attempts to identify and improve additional sources of predictability are currently underway. The following briefly describes some of the other important drivers in natural climate variability. The North Atlantic Oscillation (NAO) is a decadal signal (rather than interannual like ENSO) that affects winter weather (via precipitation) in Northern Europe and the Atlantic Coast of North America. The Pacific Decadal Oscillation (PDO) is the Pacific analogue to the NAO. Both the NAO and PDO signals tend to be secondary to ENSO in terms of intensity but do interact with ENSO, to superimpose variability on the typical weather anomalies attributed to El Niño and La Niña events. Other smaller signals also impact comparatively smaller regions of the world. For example, there appears to be a phenomenon in the equatorial Atlantic Ocean that is very similar to the Pacific ENSO. This so-called Atlantic Dipole is thought to impact weather in the Caribbean, Western Atlantic and perhaps areas in the southeastern United States such as Florida.

Forecasts of climate variability are at present based approximately equally on the products of global numerical climate models and on statistical models of the effects of ENSO. The global numerical models, often referred to as general circulation models (GCM), are relatively coarse in spatial resolution (roughly 250 km) at present due to limitations in computing power available. Currently, they are limited to the atmosphere alone, using ocean surface conditions derived from other sources. Because the variability in the ocean and the atmosphere is coupled, this approximation is inaccurate and a fully coupled model in which both the ocean and the atmosphere vary and affect each other is the ultimate goal. Furthermore, the short time scale variability (weather) of the atmosphere is not predictable beyond a few weeks, therefore climate forecasts must be probabilistic in nature. That is, they take the form of probabilities of expected anomalies. To use dynamical models in such forecasts, ensembles of realizations are required, greatly increasing the computational power needed. Statistical models based on historical observations are useful, but are restricted by the limited availability of historical data.

The present large-scale forecasts of seasonal climate anomalies appear to be useful for some purposes. However, finer spatial resolution is clearly needed in many cases. In regions where the land surface exhibits significant variability on finer scales, a process called downscaling can yield additional information. Down–scaling refers to the use of a dynamical or statistical model in conjunction with atmospheric fields predicted by a global GCM to produce fields of atmospheric winds, temperature and precipitation on a finer scale. Statistical downscaling uses historical observations to determine the relationships between the large and small scales, while dynamical downscaling uses a numerical model of the atmosphere bounded by the output of the global model to achieve the same result. The downscaling process does not produce accurate forecasts of weather events at long lead times, but has been shown in some cases to improve predictions of climate variations on small spatial scales. Therefore downscaling provides a potential means to provide information on spatial scales down to 10-20 km. However, statistical downscaling is limited by the availability of high-resolution historical data, and dynamical downscaling at present does not add useful information in all cases.

Climate forecasts provide only a portion of the necessary integrated suite of environmental forecasts that are required to make valuable predictions of health effects. Climate variability can be defined as anomalies in temperature, precipitation and other sensible weather phenomena that occur over time scales longer than those that can be predicted from initial conditions of the atmosphere. At such time scales, predictability results from the predictable effect of relatively slowly varying boundary surfaces, such as the ocean surface temperature or the land surface. Climate forecasts can be useful, for example, in allocating supplies of insecticide so as to forestall a potential vector-borne disease epidemic. At shorter time scales, less than two weeks, effective predictions of atmospheric conditions can be made from an initial state of the atmosphere by integrating a GCM forward in time. Such forecasts might be valuable for early warning of the onset of rains in a region. Finally, monitoring of current and recent events, using both remote sensing and other observations, might indicate when events that could lead to health effects have or have not occurred.

Infectious Disease Issues

In the previous report (1998) a list of disease categories (vector-, water-, food-, air-borne) was listed along with specific examples of diseases that had known or suspected relationships with weather-related variables. This list is summarized and amended in Table 2. Although there are a few additions to the previous list, it will be certain to grow as the field progresses. Perhaps of greatest importance at this early stage is to develop a set of criteria identifying and characterizing climate sensitive diseases, in general.

TABLE 2.

CLIMATE SENSITIVE DISEASES. THIS LIST INCLUDES THOSE DISEASES AND/OR PATHOGENS ORIGINALLY IDENTIFIED IN BY COLWELL AND PATZ (1998) AND NEW ADDITIONS AS PROVIDED BY PARTICIPANTS IN THE 1999 COLLOQUIUM (NOTED BY *).

Not all infectious diseases are influenced by climate or weather variability. Therefore, before embarking on any in-depth study of the relationship between climate and health, we must first define how a disease or pathogen may be characterized as “climate sensitive.” Based on our knowledge to date the following characteristics are important in this respect. First, the pathogen or vector multiplies in response to a climate or weather related variable(s) (e.g., temperature, humidity, precipitation, UV, winds, etc.). Furthermore, the survival and transport of the pathogen or vector may be related to a climate or weather related variable(s) and there is a seasonal and/or spatial pattern in the abundance of the agent. In terms of diseases themselves, cases in endemic areas are often seasonal and endemic areas of a particular disease share similar geologic and climatic characteristics. Finally, priority climate-sensitive diseases may be those for which predicting outbreaks by climatological or weather factors may substantially reduce infections (in other words, the climate component to the disease outbreak outweighs other factors in its prevalence).

In light of the above criteria, the following sections address the general climate links, disease information and modes of transmission for key disease types: vector-borne, water and foodborne, and airborne.

VECTOR-BORNE DISEASES

Vector-borne diseases were the first and most often to be associated with climate related variables. As the name suggests, vector-borne diseases are those in which a host, or vector, carries the infectious agent for part of its life history. Typical vectors are arthropods or rodents. Some of the more common diseases following this pattern are Lyme Disease and Rocky Mountain Spotted Fever (associated with ticks), dengue fever and malaria (associated with mosquitoes), among others. More recently, hantavirus was found to be a associated with the deer mouse in the southwestern United States. Additionally, certain organisms that are not considered to be vector-borne often rely on a close or symbiotic relationship where growth and multiplication is intricately linked to another organism, or host (e.g., Schistosoma with snails and Vibrio cholerae with copepods).

The relationship between the vector (or host) and climate has been noted for some time, especially for the mosquito-borne diseases, where the vector range, reproduction and biting patterns follow a distinct seasonal, geographical and climate regime.

Climate Variability and the Dynamics of Mosquito-Borne Infectious Diseases

There is a direct link between climate and the epidemiology of mosquito-borne infectious diseases (Reeves et al. 1994). This is due to both effects of climate variability on the vector population and on amplification of the pathogen within its host. Because of these well-recognized links, vector-borne diseases are among the most readily studied in terms of climate and ecological variables.

Infected mosquitoes transmit some pathogens, such as malaria and dengue fever, directly from person to person. Other diseases, such as St. Louis Encephalitis and West Nile fever, are transmitted to humans only incidentally; the natural life cycle is between birds and mosquitoes. The prevalence of mosquito-borne pathogens is monitored either by looking for human infection or for infection within vertebrate reservoir hosts (i.e., birds) and/or mosquitoes. In human populations, infection and disease generally follow a predictable time course similar to the curve depicted in Figure 1. The effective window for early intervention is the time interval between initial detection, when the pathogen has amplified above a minimum threshold level in natural hosts, and infection in humans. The objective of monitoring is to detect pathogen amplification at the earliest possible time to prevent epidemics. By using climate data, the sensitivity of surveillance is improved by essentially lowering the threshold of detection and increasing the length of the response (intervention) time. Temperature driven factors, such as mosquito population growth and pathogen replication rates, allow temperature to be used as a proxy for early detection (this is seen in the slope of the curve in Figure 1). The height of the curve, which represents the maximum level of the pathogen, is related to mosquito population size, which in turn is dependent upon water availability for larval habitat. Therefore, both temperature and precipitation can be used to model mosquito vector populations and to predict periods of human susceptibility.

FIGURE 1.

VECTOR-BORNE DISEASE TRANSMISSION CURVE SHOWING HOW THE THRESHOLDS OF DETECTION AND REPORTING OF HUMAN INFECTION DELINEATE THE RESPONSE TIME FOR INTERVENTION. CLIMATE MODELS MAY BE USED TO FORECAST THE RISK OF DISEASE EMERGENCE AND ALLOW EARLY WARNING (more...)

A complete monitoring program for a mosquito-borne disease should contain the parameters shown in Table 3 (Eldridge 1987), which includes climate projections from global models and local measurements by weather stations. Landscape change can be detected by remote sensing. Vector population size can be measured by sampling mosquitoes in the environment. Sampling methods are mosquito species specific; some species are captured in traps, whereas others are captured as they come to feed on vertebrate hosts or human bait. Additionally, the vector population is often tested to determine the percent infected and is a better estimate of transmission risk than simple vector abundance. In the case of those diseases with a non-human vertebrate reservoir, the prevalence of infection in the wild animal can be determined, or alternatively sentinel vertebrate hosts can be tested to see when they become infected. Finally, reporting of human cases can be used for monitoring if the situation becomes epidemic.

TABLE 3.

NESTED COMPONENTS OF A VECTOR-BORNE DISEASE MONITORING PROGRAM COMBINING LONG RANGE CLIMATE AND SHORT RANGE WEATHER FORECASTS, LANDSCAPE CHANGE, VECTOR ABUNDANCE AND INFECTION, RESERVOIR HOST DISEASE [ZOONOSES ONLY], AND HUMAN CASES.

The objective of monitoring is to provide sufficient early warning so that intervention measures, such as mosquito control, immunization of the at-risk population, or public notification, can be initiated to prevent human illness. Because these measures require a response time, the use of climate forecasts is aimed at projecting infection at an earlier date than existing monitoring methods (Linthicum et al. 1999). The close range validation of these predictive models is achieved by monitoring weather and landscape changes (Pope et al. 1992).

While short-term predictions allow responding agencies enhanced lead-time to anticipate pending outbreaks, long-range climate predictors can be used to alert the control agencies to the possibility of future problems. Near time predictors will initiate an escalating cascade of control activities that can range from increased searching and control of mosquito larval sources to stockpiling of vaccine. When the agent is detected, adult mosquito control is initiated to interrupt transmission and drugs and vaccines are administered, if applicable, to the local community. When multiple human cases occur, an epidemic is in progress, and the above measures are intensified and expanded over a wider area.

St. Louis Encephalitis (SLE) in the western United States provides an excellent example of the application of a multi-faceted surveillance program that integrates information on climate variability, vector abundance and infection, and infection rates in the vertebrate reservoir host population (Eldridge 1987, Reisen 1995). Figure 2 shows a diagrammatic presentation of the cascade of events leading to an outbreak such as occurred in Bakersfield in 1989 (Reisen et al. 1992). In California, the vector, Culex tarsalis, terminates hibernation in January. The females immediately take a blood meal and search for sites to lay eggs. The number of potential larval habitats is related to the amount of winter rain, which creates surface pools. The rate of development of the first and subsequent generations is dependent upon temperature. The size of the summer population is dependent upon the amount of surface water larval habitats created by run-off of snowmelt in the Sierra Nevadas and by agricultural irrigation in the mid-summer. The appearance of virus is dependent on temperatures that must exceed a minimum threshold of 16˚ C for replication in the mosquito host (Reisen, et al. 1993). When the mosquito population size exceeds minimum thresholds (Olsen et al. 1979), virus activity can be detected by isolation from pools of mosquitoes or by serological conversions (presence of antibody) in wild or sentinel bird populations. After amplification exceeds a minimum threshold, transmission to humans occurs. Establishing firm links between climate projections, snow pack, and river run-off may allow the accurate forecasting of mosquito population size for the following summer and perhaps the risk of virus transmission (Wegbreit and Reisen 2000).

FIGURE 2.

SEQUENCE OF EVENTS DURING A HYPOTHETICAL OUTBREAK OF A MODEL MOSQUITO TRANSMITTED ZOONOSIS, ST. LOUIS ENCEPHALITIS. DATA SHOWN INCLUDE VECTOR MOSQUITO ABUNDANCE [AVG. CULEX TARSALIS PER TRAP], VIRUS ISOLATIONS FROM POOLS OF MOSQUITOES, BIRD RESERVOIR (more...)

Recommendations

Research is needed to enable improved prediction of outbreaks and implementation of early prevention aids. Specifically, the relationships for individual diseases, infectious agents and vectors are unique. Therefore, efforts must be taken to establish baseline levels of vector geographic range and abundance, and how these could change under different climate variability scenarios. Furthermore, patterns should be identified in disease endemic areas. For instance: do outbreaks follow a particular sequence of climate factors (i.e., flood followed by drought, etc.), for a mosquito vector is flooding in the form of sea level rise more important than precipitation, and when do mosquitoes tend to feed/bite? Finally, potential areas of human adaptation must be identified, particularly when human activities or modifications overshadow an otherwise clear link between climate and the disease (i.e., air conditioning).

WATER- AND FOOD-BORNE DISEASES

The quality of water and food also can be related to climate and weather variability and includes drinking and irrigation waters as well as both fresh and marine surface waters. For example, oceanic and coastal waters are known to harbor and transport microorganisms that cause disease in humans and other animals. As modulators of climate, oceans also indirectly influence disease patterns and distribution of many pathogens. While certain pathogenic or toxigenic microorganisms, including toxic phytoplankton and Vibrio spp., occur naturally in marine and estuarine waters, anthropogenic contaminants including enteric bacteria, protozoa and viruses may be introduced to coastal waters as sewage pollution. Despite the relatively unfavorable environment, these introduced organisms may survive for prolonged periods in the marine environment, often associated with sediments and other protective environments (LaBelle and Gerba 1982).

Interannual variability associated with signals such as the El Niño-Southern Oscillation (ENSO) has been proposed to influence cholera outbreaks in Perú (Colwell 1996) and has been associated with dynamics of the disease in Bangladesh (Pascual et al. 2000), and the levels of anthropogenic pollution (including human viruses) in estuaries of south Florida (Fig. 3) (Lipp et al. in press a, Lipp et al. in press b). Furthermore, ENSO has been associated with levels of diarrheal disease in Perú (Checkley et al. 2000). In addition to these periodic climate “anomalies,” a signal in long-term sea surface temperature records recently has been identified and is believed to be a manifestation of global warming (Livezey and Smith 1999). The impact of long-term climate changes will likely include shifts in the distribution of many species, including pathogenic microorganisms, and may alter host susceptibility to infection. For example, the geographic range of many Vibrio spp. would be expected to increase with warming temperatures. Furthermore, changes may affect toxicity in certain organisms such as Pfiesteria piscicida in which the toxic zoospore stage is thought to be secondarily related to warm temperatures (the presence of fish is the primary signal for transformation to toxic stage) (Burkholder et al. 1998).

FIGURE 3.

THE POTENTIAL ROLE OF THE 1997-1998 EL NIÑO EVENT ON WATER QUALITY IN SOUTHWEST FLORIDA. AS THE EL NIÑO PEAKS IN THE WINTER MONTHS (SHOWN BY NIÑO 3.4 MONTHLY SEA SURFACE TEMPERATURE ANOMALIES), PRECIPITATION IN SOUTHWEST FLORIDA (more...)

Rainfall and runoff have been associated with individual outbreaks of waterborne disease and it is clear that pathogens of fecal origin can find their way into water. Until recently the importance of rainfall in association with the transport of contaminants from their sources to drinking water had not been studied on a watershed basis. An investigation in the U.S. has found that between 20 and 40% of the surface water and ground water outbreaks from 1971 to 1994 were associated with extreme precipitation. This relationship was statistically significant and spatial clustering of outbreaks in key watersheds demonstrated the importance of sources and risk during rainfall events (Rose et al. 2000). With climate predictions suggesting that storms will be of greater intensity and that the average precipitation event is likely to be heavier, the risk of waterborne disease in the U.S. is likely to increase.

As is the case for water, human pathogens in food generally have two kinds of origin: anthropogenic or naturally-occurring. In fish or any other seafood, pathogens may occur naturally in the organisms’ native habitat (i.e., Vibrio spp.). Food products, including seafood, may also be contaminated with enteric pathogens during handling while fruits and vegetables may become contaminated with poorly treated irrigation waters. For enteric pathogens, fecal indicator species are often used to determine the level of contamination but are ineffective as a proxy for naturally-occurring pathogens, which may be more susceptible to climate variability. Therefore, by monitoring environmental parameters, a global climate model for specific pathogens would be useful for control and prevention. Unfortunately, not much information is available. Methods are needed to identify emerging and specific pathogens to determine their persistence, survival and distribution in the environment in relation to climate and human heath.

For both food and water, local weather patterns may play a role in the dispersal of pathogenic microorganisms. Events such as extreme rainfall and floods often over-burden water treatment facilities and increase storm water run-off; both may result in the introduction of high levels of enteric pathogens to nearby surface waters and wells, if groundwater becomes contaminated. Heavy rainfall and associated flooding have been statistically related to high incidence of gastrointestinal disease (Gueri et al. 1986). In coastal regions disease often occurs through recreational exposure or consumption of contaminated fish or shellfish (Fleisher et al. 1996, Wittman and Flick 1995). Furthermore, agents that cause disease from a small dose are even more problematic (i.e., viruses). Weather patterns may also affect the distribution and ecology of naturally-occurring disease agents such as Vibrio cholerae and V. vulnificus.

Modeling of cholera dynamics as a function of climate variability

The modeling of cholera in relation to climate is at the first stage of identifying the key environmental and climate parameters associated with disease variability. As for many diseases, in endemic regions such as Bangladesh, the incidence of cholera cases follows seasonal patterns. Further analysis of such patterns reveals a relationship to climate at both seasonal and interannual time scales. Recent work by Pascual et al. (2000) provides time series evidence for the role of climate in the interannual variability of cholera that had been previously hypothesized.

Environmental parameters such as temperature have been implicated in the seasonal distribution of toxigenic Vibrio cholerae in the environment and the occurrence of clinical cases of cholera during certain times of the year (Huq et al. 1990, Colwell 1996). Lobitz et al. (2000) show that sea surface temperatures and sea surface height are directly correlated with cholera epidemics in Bangladesh. Preliminary evidence shows a similar trend in Perú. In addition, other climate related variables are being observed closely for use in development of predictive models of cholera. Both statistical and mathematical models would be appropriate given that the sources of variability are often numerous and not well defined. Clearly, there is a need for identifying the environmental and climatological parameters mediating a local or regional effect of ENSO on cholera. Candidates include water temperature, rainfall, salinity and the abundance of plankton populations. In addition, there is a need for further analyses of cholera time series spanning multiple ENSO events.

Future model development will benefit from on-going research on environmental and climatological influences on the population dynamics of pathogens, such as Vibrio cholerae in aquatic environments (Lobitz et al. 2000). Key issues will be to understand the connection between pathogen abundance in environmental reservoirs and disease dynamics (Franco et al. 1997), the feedback from the human population to the aquatic reservoirs, and the interplay of climate parameters with the determinants of disease susceptibility in the human population.

Recommendations

The complex relationships between climate, weather, water, food and specific pathogens must be better defined. Currently, the level of understanding of the linkages between water or foodborne disease and climate is less sophisticated than that for vector-borne diseases.

Historical assessments offer the best currently available tools to elucidate important relationships. Results can then guide prospective research emphases, which will define those climate or weather variables important to pathogen prevalence in the aquatic environment and suggest the best geographical scale for investigation. Finally, long term investigations incorporating surveillance for both pathogens/diseases and environmental parameters using sophisticated tools (i.e., molecular) are required.

AIR-BORNE DISEASES

The common “cold” and “flu” are probably diseases that would be most likely characterized as airborne. The primary microorganisms associated with these conditions are viruses, Rhinovirus and Influenza. Climatic factors have been suspected to contribute partly to the seasonal cycles that are very apparent for Influenza, with the peaks in disease occurring in the northern hemisphere between mid-December and February. The relationship is not clear-cut, however, as mean monthly temperature is not the driving force (Langford and Bentham 1995). Exposure studies in mice have suggested that humidity may play a significant role in the survival of the virus in droplets that are aerosolized from infected individuals (Schulman and Kilbourne 1963). Climatic factors may also influence intermediate-host bird populations and geographical distribution of the disease. However, these factors have not been fully explored or investigated from a scientific vantage that attempts to determine the climatic attributes that may be most important.

Rhinoviruses survive on surfaces and hands and are spread through contact. While low temperature will increase the virus survival there is no evidence that this is connected to the ambient climatic conditions but more to the microenvironmental conditions.

The field of airborne allergens can also be connected and studied in conjunction with seasonal and climatic factors. While there is already a seasonal nature to flowering and pollination as hay fever sufferers know well, the geographic range of particular species and duration of pollen season may be altered by a changing climate. Temperature and rainfall would likely be the most important environmental factors, along with measurement of outcomes such as the occurrence and severity of asthma (Ahlholm et al. 1998).

Airborne agents also affect marine resources. Remote sensing data have been used by scientists at the USGS to follow dust from the Sahara as it makes it way across the Atlantic Ocean, ultimately influencing the Caribbean and Gulf of Mexico. Sampling of this dust has shown that it contains higher levels of microorganisms compared to the ambient air and pathogenic species have been identified (D.W. Griffin unpublished data). Illnesses thought to be associated with Saharan dust have been identified in areas such as Puerto Rico, but to date firm scientific evidence linking these events is lacking. Season, wind patterns and humidity are likely involved in this phenomenon

Recommendations

Programs should be initiated to begin to explore how climate influences air-borne pathogens, given that little data are presently available. Everything from the global influence of Saharan dust to the backyard source of airborne allergens to the seasonal spread of the flu should be included in future programs.

DISEASE SURVEILLANCE

One of the key factors to better study and understand the relationship between health and climate is the availability of long term disease and pathogen data. Unfortunately, throughout the developed world systematic disease surveillance is being abandoned (including the United States). Therefore, while adequate historical databases are often difficult to obtain, there will soon be a lack of new consistent health statistics. At a time when researchers need to find data at central locations, more and more they will be forced to obtain data from local sources, which may be inconsistent and difficult to compare between sources. To advance our understanding of the relationships between climate variability and infectious diseases, the following data are needed. Long term, historical disease and pathogen surveillance data must be located, salvaged, and electronically archived. Continuing emphasis should be placed on prospective long-term surveillance that includes medical, ecological, and climatological parameters at periods greater than five years.

In addition to the lack of disease surveillance data there is an even greater paucity of long term surveillance for key pathogens in the environment. To date there are few long-term monitoring efforts that are supported and currently, we must rely on short term locally focused studies or a few inconsistent longer term data sets. More emphasis needs to be placed on surveying the environment for the pathogen and/or vector of concern (i.e., before there is disease and reporting to local health officials). This will allow us to begin to emphasize environmental conditions that contribute directly or indirectly to cause an epidemic of a specific disease, merging the fields of epidemiology and ecology.

Molecular methods for environmental surveillance provide very powerful tools to examine and better understand shifts and recognize novel strains of microorganisms. Such was the case in the study of the West Nile Virus following its dramatic introduction to New York. Such studies on the genetic diversity of the virus throughout the world were and will be useful in helping to answer many questions. Is this a new strain of virus? Where did it come from? Will it establish itself in animal reservoirs? Will it move geographically? How fast will the virus spread? Studies on the virus, ecology of the vector, environmental conditions and climate factors are imperative. This will lead to the ability to predict potential outbreaks based on susceptible populations, regions, and climate variability. This is especially important in developing nations where limited resources may be more efficiently allocated in advance of a health crisis. Clearly the goal is to take health, environment and climate data and use these in understanding and incorporating climate signals, as proactive, rather than reactive, approaches for protection of public health.

Recommendation

Disease surveillance and environmental monitoring for microorganisms of interest using conventional and molecular tools within the context of ecological studies should be combined with gathering of climatic and weather data at a minimum of five-year periods.

Tools, Information and Models for Coupling Climate and Health

IN ORDER TO ADVANCE THE SCIENCE OF CLIMATE AND DISEASE, WE MUST INCREASINGLY RELY ON MODELS. BECAUSE OF THE LONG TIME SERIES REQUIRED TO IDENTIFY PATTERNS AND THE DIFFICULTY IN ASSESSING CLIMATE VARIABILITY ON PATHOGENS IN THE LABORATORY, MODELS WILL PLAY AN IMPORTANT ROLE IN THE UNDERSTANDING AND PREDICTION OF DISEASE OUTCOMES UNDER PARTICULAR CLIMATE SCENARIOS. MODEL VALIDATION WILL REQUIRE IN DEPTH RETROSPECTIVE ANALYSES OF AVAILABLE DATABASES AND PROSPECTIVE ANALYSES OF ONGOING SURVEILLANCE.

Collecting and Managing Appropriate Climatological, Ecological and Health Data

One of the obstacles in the development of a successful integrated and multidisciplinary research effort will be data management and acquisition. Inherently, projects falling under this interdisciplinary research umbrella will require very large and long-term data sets with many parameters. Therefore a key question is, how do we collect and manage appropriate climatological, ecological and health data?

To address this topic, the scientific community should know which data are appropriate. We are only beginning to compile a general list and more items will be included as the field progresses. Table 4 lists the data needs known date and is obviously not exhaustive. Additionally, existing key databases must be located and accessed. Table 5 (available soon in amended report) lists some databases identified to date, which agencies administer them and how they may be or have been applied to climate and disease research.

TABLE 4.

CLIMATOLOGICAL, ECOLOGICAL AND HEALTH DATA NEEDS FOR INTERDISCIPLINARY STUDY.

Recommendations

The above obstacles can begin to be solved by first engaging in data mining. Key databases should be identified in all sectors and progress made towards accessibility. Public agencies should be encouraged to continue surveillance programs. Long term studies in both health and ecology are a priority to incorporate all important parameters and meta-data in mechanistic and semi-mechanistic models. Finally, a public agency should serve as an archival point for all key data sets. This storage system would ideally be web-accessible with appropriate pass codes and security measures put in place.

AN EMERGING SCIENCE: BIOCLIMATOLOGY

IN THE LAST THREE YEARS, SCIENTISTS IN THE FIELDS OF CLIMATOLOGY, METEOROLOGY, MICROBIOLOGY, MEDICINE, MEDICAL ENTOMOLOGY, ECOLOGY, EPIDEMIOLOGY, OCEANOGRAPHY, AND SPACE SCIENCE HAVE COLLABORATED TO STUDY HOW NATURAL CLIMATE VARIABILITY AFFECTS OCCURRENCE AND PREVALENCE OF PATHOGENIC MICROORGANISMS, VECTORS, AND DISEASE OUTCOMES. HOWEVER, TO DO THIS, SCIENTISTS HAVE SPENT MUCH TIME FINDING A COMMON “LANGUAGE,” TRAINING EACH OTHER IN THE TERMINOLOGY USED, TYPES OF DATA AVAILABLE AND INTERPRETATION OF SUCH data as well as the tools available for analysis. Inadvertently, the field of “bioclimatology” has emerged and younger scientists now have an opportunity for a multidisciplinary education. Clearly, there is a need to support such interdisciplinary training in “bioclimatology” and this should be encouraged at all levels—including development of undergraduate courses, graduate programs that begin to merge the fields and postdoctoral training grants. Increased communication between disciplines is imperative. There will be a need for meetings and journals that focus on climate issues with biological and medical applications. Scientific societies may be best able to accomplish this goal.

General Recommendations

THE FOLLOWING LIST SUMMARIZES RECOMMENDATIONS FROM INDIVIDUAL SECTIONS WITHIN THE REPORT.

• Climatologists, health scientists and ecologists must find common ground for successful research and engage in discussions that allow the others to know which data are needed and available.
• Long-term ecological surveillance for pathogens and vectors in conjunction with appropriate collection of disease and climate data is imperative. While data are needed for all disease types (vector, water, food and air borne), we are currently severely lacking in research relating airborne disease to climate.
• Surveillance programs should take advantage of more sophisticated molecular tools.
• Historical data sets must be identified, electronically stored, archived and made accessible.
• Studies should move towards predictive models and integrated risk assessment frameworks.
• Research should begin to incorporate remotely sensed data to track and predict disease outbreaks and high levels of key pathogens.
• Institutional support for “bioclimatology” training should be encouraged.

View in own window

	THIS SIGN HAS BEEN USED FOR AIR IN OLD CHEMISTRY.
	ONE OF THE METEOROLOGICAL SIGNS FOR HAIL, SOFT HAIL.
	THIS SIGN HAS BEEN USED IN SOME METEOROLOGICAL SYSTEMS FOR HIGH DRIVING SNOW OR A KIND OF LIGHT SNOWSTORM.
	THE CURVED ARROW IS AN EXTREMELY OLD SIGN FOUND ON THE WALLS OF PREHISTORIC CAVES IN WESTERN EUROPE. TODAY, IS USED AS A METEOROLOGICAL AND CARTOGRAPHIC SIGN INDICATING CONSTANT WINDS.
	A SIGN USED IN SOME ENGLISH METEOROLOGICAL SYSTEMS TO DENOTE GROUND FROZEN HARD AND DRY.
	BOTH THE SQUARE AND THE CROSS WITH EQUAL ARMS ARE RELATED TO THE EARTH, THE GROUND, AND THE LAND. IN CHINESE AND JAPANESE IDEOGRAPHY, THIS ENTRY SIGN ACTUALLY MEANS FIELD OR GROUND.
	THIS SIGN MEANS LOW DRIVING SNOW.
	THIS SIGN HAS BEEN USED IN METEOROLOGY TO DENOTE WINDS OF VARYING STRENGTHS. THE STRENGTH OF THE WIND IS INDICATED BY THE NUMBER OF SHORT, PERPENDICULAR LINES ACCORDING TO A SIMPLE CODE.
	A METEOROLOGICAL SIGN USED IN THE UNITED STATES TO INDICATE HURRICANE OR TORNADO.
	IN METEOROLOGY, THIS SIGN HAS BEEN OR IS USED TO INDICATE ICE OR A COATING OF ICE.
	THIS SIGN HAS BEEN USED IN SOME METEOROLOGICAL SYSTEMS TO SIGNIFY A DAMP MIST.
	THIS SIGN HAS BEEN USED IN ENGLISH METEOROLOGY TO DENOTE GROUND PARTLY COVERED BY SNOW.
	USED IN ALCHEMY TO DENOTE THE PROCESS BY WHICH A VAPORIZED SUBSTANCE IS TRANSFORMED INTO A LIQUID FORM: A PRECIPITATION. THIS SIGN HAS ALSO BEEN USED TO DENOTE THE RESULT OF SUCH A PROCESS, THE PRECIPITATE.
	THE WAVY ARROW IS A COMMON SIGN FOR SEA CURRENTS ON MAPS AND NAUTICAL CHARTS, ETC.
	A METEOROLOGICAL SIGN FOR SNOW SHOWERS, SQUALLS.
	SOURCE FOR SYMBOLS AND THEIR DEFINITIONS: LIUNGMAN, CARL G. 1991. DICTIONARY OF SYMBOLS. ABC-CLIO, INC. SANTA BARBARA, CA.

BIBLIOGRAPHY

• Ahlholm, J.U., M.L. Helander, J. Savolainene. 1998. Genetic and environ mental factors affecting the allergenicity of birch (Betula pubescens czerepanovii [Orl.] Hamet-ahti) pollen. Clin Exp Allergy 28:28-1384. [PubMed: 9824411]
• Burkholder, J.M., H.B. Glasgow Jr., and A.J. Lewitus. 1998. Physiological ecology of Pfiesteria piscicida with general comments of “ambush-predator” dinoflagellates. In Physiological Ecology of Harmful Algal Blooms, D.M. Anderson et al., eds. NATO ASI Series Vol. G 41. Springer-Verlag, Berlin. Pp. 175 – 191.
• Checkley, W. L.D. Epstein, R.H. Gilman, D. Figueroa, R.I. Cama, J.A. Patz and R.E. Black. 2000. Effect of El Niño and ambient temperature on hospital admissions and diarrhoeal diseases in Peruvian children. Lancet 355: 442-450. [PubMed: 10841124]
• Colwell, R.R. 1996. Global climate and infectious disease: the cholera paradigm. Science 274: 2025-2031. [PubMed: 8953025]
• Colwell, R.R. and J.A. Patz. 1998. Climate, Infectious Disease and Health: An Interdisciplinary Perspective. American Academy of Microbiology, Washington, DC. [PubMed: 32687286]
• Eldridge, B.F. 1987. Strategies for surveillance, prevention and control of arbovirus diseases in western North America. American Journal of Tropical Medicine and Hygiene 37: 77S-86S. [PubMed: 2891313]
• Ellner, S.P., B.A. Bailey, G.V. Bobashev, A.R. Gallant, B.T. Grenfell and D.W. Nychka. 1998. Noise and nonlinearity in measles epidemics: combining mechanistic and statistical approaches to population modeling. The American Naturalist 151: 425-440. [PubMed: 18811317]
• Fleisher, J.M., D. Kay, M. Wyer and H. Merrett. 1996. The enterovirus test in the assessment of recreational water-associated gastroenteritis. Water Research 30: 2341-2346.
• Franco, A.A., A.D. Fix, A. Prada, E. Paredes, J.C. Palomino, A.C. Wright, J.A. Johnson, R. McCarter, H. Guerra and J.G. Morris. 1997. Cholera in Lima, Peru correlates with prior isolation of Vibrio cholerae from the environment. American Journal of Epidemiology 146: 1067-1075. [PubMed: 9420531]
• Gueri, M., Gonzalez, C., and Morin, V. 1986 The effect of the floods caused by “El Niño” on health. Disasters 10: 118–124.
• Haas, CH., J.B. Rose, and C.P. Gerba (eds). 1999. Quantitative Microbial Risk Assessment. John Wiley and Sons, NY, NY.
• Huq, A., R.R. Colwell, R. Rahman, A.A. Chowdhury, S. Parveen, D.A. Sack and E. Russek-Cohen. 1990. Detection of Vibrio cholerae O1 in the aquatic environment by fluorescent-monoclonal antibody and culture methods. Appl Environ Microbiol. 56: 2370-2373. [PMC free article: PMC184735] [PubMed: 2206100]
• LaBelle, R. and C.P. Gerba. 1981. Investigations into the protective effect of estuarine sediment on virus survival. Water Research 16: 469-478.
• Langford, I.H. and G. Bentham. 1995. The potential effects of climate change on winter mortality in England and Wales. Int J Biometeorol. 38: 141-147. [PubMed: 7744529]
• Linthicum, K.J., A. Anyamba, C.J. Tucker, P.W. Kelley, M.F. Myers, and C.J. Peters. 1999. Climate and satellite indicators to forecast Rift Valley fever epidemics in Kenya. Science. 285: 397-400. [PubMed: 10411500]
• Lipp, E.K., R. Kurz, R. Vincent and C. Rodriguez-Palacios, S.R. Farrah and J.B. Rose. In Press a. The effects of seasonal variability and weather on microbial fecal pollution and enteric pathogens in a subtropical estuary. Estuaries.
• Lipp, E.K., N. Schmidt, M. Luther and J.B. Rose. In Press b. Determining the effects of El Niño-Southern Oscillation events on coastal water quality. Estuaries.
• Livezey, R.E. and Smith, T.M. 1999. Covariability of aspects of North American climate with global sea surface temperatures on interannual to interdecadal timescales. Journal of Climate 12: 289 – 302.
• Lobitz, B.L., L Beck, A. Huq, B. Wood, G. Fuchs, A.S.G. Faruque and R.R. Colwell. 2000. Climate and Infectious Disease: Use of remote sensing for detection of V. cholerae by indirect measurement. Proc. Natl. Acad. Sc. 97:97-1438. [PMC free article: PMC26452] [PubMed: 10677480]
• Olsen, J.G., W.C. Reeves, R.W. Emmons, and M.M. Milby. 1979. Correlation of Culex tarsalis indices with the indices of St. Louis encephalitis and western equine encephalomyelitis in California. American Journal of Tropical Medicine and Hygiene 28: 335-343. [PubMed: 453436]
• Pascual, M., X. Rodó, S.P. Ellner, R. Colwell, and M.J. Bouma. 2000. Cholera Dynamics and El Niño-Southern Oscillation. Science 289: 1766-1769. [PubMed: 10976073]
• Pope, K.O., E.J. Sheffner, K.J. Linthicum, C.L. Bailey, T.M. Logan, E.S. Kasischke, K. Birney, A.R. Njogu and C.R. Roberts. 1992. Identification of central Kenyan Rift Valley Fever virus vector habitats with Landsat TM and evaluation of their flooding status with airborne imaging radar. Remote Sensing of the Environment 40: 185-196.
• Reeves, W.C., J.L. Hardy, W.K. Reisen, and M.M. Milby. 1994. Potential effect of global warming on mosquito-borne arboviruses. Journal of Medical Entomology 31: 323-332. [PubMed: 8057305]
• Reisen, W.K. 1995. Guidelines for surveillance and control of arboviral encephalitis in California: 1-34. Intragency Guidelines for the Surveillance and Control of Selected Vector-Borne Pathogens in California. Sacramento, CA. Mosquito Vector Control Association of California.
• Reisen, W.K., R.P. Meyer, M.M. Milby, S.B. Presser, R.W. Emmons, J.L. Hardy, and W.C. Reeves. 1992. Ecological observations on the 1989 outbreak of St. Louis encephalitis virus in the southern San Joaquin Valley of California. Journal of Medical Entomology 29:29-472. [PubMed: 1625296]
• Reisen, W.K., R.P. Meyer, S.B. Presser, and J.L. Hardy. 1993. Effect of temperature on the transmission of western equine encephalomyelitis and St. Louis encephalitis viruses by Culex tarsalis (Diptera: Culicidae). Journal of Medical Entomology 30:30-151. [PubMed: 8433322]
• Rose, J.B., S. Daeschner, D.R. Easterling, F.C. Curriero, S. Lele, and J.A. Patz. 2000. Climate and Waterborne disease outbreaks. J. Amer. Water Works Assoc. 92: 77-87.
• Schmidt, N., E.K. Lipp, J.B. Rose and M. Luther. 2001. Analysis of ENSO related trends in Florida precipitation and streamflow. Journal of Climate 14: 615-628.
• Schulman, J.L. and E.D. Kilbourne. 1963. Experimental transmission of Influenza virus infection in mice: some factors affecting the incidence of transmitted infection. J.Exp Med. 118:118-267. [PMC free article: PMC2137714] [PubMed: 14074390]
• Wegbreit, J. and W.K. Reisen. 2000. Relationships among weather, mosquito abundance and encephalitis virus activity in California: Kern County 1990 – 1998. Journal of the American Mosquito Control Association 16: 22-27. [PubMed: 10757487]
• Wittman, R.J. and G.J. Flick. 1995. Microbial contamination of shellfish: Prevalence, risk to human health and control strategies. Annual Review of Public Health 16: 123–140 [PubMed: 7639867]