1. Introduction
Cholera is a waterborne disease (WBD) endemic to the Bangladesh region and responsible for numerous global pandemics throughout history, including the current era (Kohn 1995). Infection results from ingesting water contaminated with the bacteria Vibrio cholerae, and, when untreated, mortality rates as high as 50% are common. With proper treatment, mortality rates drop, instead, to less than 1% even in severe cases (Cook 1996). However, in poor countries with limited infrastructure, such as Bangladesh, it is not always possible to provide treatment as quickly as is necessary once an outbreak begins. Hence, a means of forecasting heightened cholera risk would be of great benefit to society.
Cholera incidence in Bangladesh follows a distinct annual cycle. On average, cases peak during the Northern Hemisphere (NH) spring, prior to the arrival of the monsoon rains, and again in September–October after the rains retreat (Glass et al. 1982). This clear annual cycle, and its apparent relationship to the monsoon, suggests that there may be an environmental component to cholera incidence in this region. The hypothesis that Vibrio cholerae concentration and hence cholera incidence is influenced by environmental factors such as temperature, salinity, and pH is supported by numerous studies (Colwell 1996; Franco et al. 1997; Pascual et al. 2000, 2002; Rodó et al. 2002; Koelle and Pascual 2004; Koelle et al. 2005a, b), although a full understanding of the nature of this influence is still a subject of debate.
Of particular interest is the existence of a correlation between interannual cholera variability, independent of internal disease dynamics, and the Niño-3.4 index (Koelle et al. 2005b). Maximum correlations occur when the Niño-3.4 index, one measure of the El Niño–Southern Oscillation (ENSO), leads cholera by 11 months. This relationship has been shown, in conjunction with a model of the internal dynamics of the disease–host relationship, to improve forecasts of cholera risk over the results of a disease model alone (Pascual et al. 2008). However, our understanding of this relationship is incomplete in the absence of a physical mechanism to link tropical Pacific sea surface temperature (SST) to cholera variability in a region as far removed as Bangladesh. It is always possible that the correlation with the Niño-3.4 index is purely coincidental, and this possibility must be explored if the relationship is to be of practical value.
Despite the extensive literature describing interactions between ENSO and the monsoon (e.g., Walker 1923, 1924; Rasmusson and Carpenter 1983; Goswami 1998; Ju and Slingo 1995; Krishnamurthy and Goswami 2000; Krishnamurthy and Shukla 2000), the impact of ENSO on rainfall in Bangladesh has received relatively little attention. The vast majority of monsoon studies focus solely on rainfall over India (e.g., Goswami et al. 1999; Krishnamurthy and Shukla 2000), which is poorly correlated with rainfall over Bangladesh. While there are studies that focus specifically on Bangladesh (e.g., Chowdhury 2003), these studies tend to focus on contemporaneous correlations spanning the period from 1948 to the present. This presents difficulties for our study of climate links to cholera for several reasons. The dominant strain of cholera in the environment changed during the first part of this period: the so-called classical strain was replaced in the latter part of the period by the El Tor strain, possibly due in part to shifts in climate (Koelle et al. 2005a), and the timing of peaks in these two strains is different. The relationship between ENSO and the monsoon is also known to have changed in the mid-1970s. Thus, relationships derived from considering all events from 1948 forward, as was done in Chowdhury (2003), may combine events with different dynamical features and includes numerous events outside of our period of interest.
One plausible mechanism by which ENSO could affect cholera in Bangladesh is through alterations in the monsoon circulation. It is well established that ENSO interacts with the monsoon through the so-called atmospheric bridge (e.g., Klein et al. 1999; Alexander et al. 2002). Anomalies in tropical Pacific SST lead to shifts in the ascending and descending branches of the Walker circulation, which in turn influence the timing and intensity of the monsoon circulation and rainfall. In this work we investigate potential links between the climate of the Bangladesh region and tropical eastern and central Pacific SST, with the goal of providing a dynamical mechanism to explain the apparent influence of ENSO in the preceding winter on fall cholera incidence (Koelle et al. 2005a, b). Our primary tool in this study is a regionally coupled model, which is described in detail in section 2. Results from the analysis are described in section 3, and a summary of our results and conclusions are presented in section 4.
2. Data and methodology
The results presented here are from a “pacemaker” or regionally coupled model. In this methodology, SST is prescribed as a lower boundary condition for a dynamical global atmospheric general circulation model (AGCM) in a portion of the domain and the AGCM is typically coupled to a mixed layer outside of that domain (e.g., Alexander 1992a, b; Alexander et al. 2002; Bladé 1997, 1999; Lau and Nath 2000, 2003; Shinoda et al. 2004; Wu and Kirtman 2004a, b). The term pacemaker refers to the consistent signal provided by the prescribed region. The large amplitude pattern of SST anomalies associated with ENSO in the eastern tropical Pacific drives circulation and rainfall anomalies worldwide, thus specifying SST in that region has special significance. We have adopted this methodology for this study for two reasons. First, inconsistencies in the surface energy budget that typically occur in AGCM simulations with prescribed SST, particularly in the Indian Ocean, can have a negative impact on the simulation of the monsoon (e.g., Wu and Kirtman 2004a). Coupling the AGCM to a mixed layer ocean outside of the pacemaker region allows for a more consistent surface energy budget, improving the accuracy of the simulation of the monsoon over a model with fully prescribed SST. Second, the model output must be directly comparable to the observed record of cholera incidence. This criterion prevents us from using a fully coupled model, as coupled models quickly diverge from the observed climate record. By prescribing SST in the tropical Pacific, we ensure that ENSO events with the same intensity and timing as the observed are present in our simulations and that we can directly compare the results from the model to the observations insofar as ENSO is responsible for the observed behavior. The eastern tropical Pacific has also been identified as a region where the ocean primarily forces the atmosphere rather than the atmosphere forcing the ocean (Wu et al. 2006), and it is well separated from our region of interest. Thus we feel the pacemaker methodology is uniquely suited to examining the influence of Pacific SST on cholera in India and Bangladesh. The experimental design allows for energetically consistent atmosphere–ocean feedbacks in the Indian Ocean and Bay of Bengal, while at the same time incorporating the observed record of Pacific SST in the regions believed to influence cholera variability.
The model used in this study is the Center for Ocean-Land-Atmosphere Studies (COLA) AGCM, version 3.1 (v3.1). This is an update of the COLA AGCM, v2.2, which has been used extensively for modeling studies in the past (e.g., Schneider 2002; Kirtman and Shukla 2002; Kirtman et al. 2002; Kirtman 2003), and is closely related to the COLA, v3.2, model described by Misra et al. (2006). The model has 28 vertical levels and is run at T62 resolution. As described above, the ocean domain is separated into two regions. In the tropical Pacific (eastern and central) and polar regions (see Fig. 1a), we prescribe observed monthly-mean SST from the Hadley Centre Sea Ice and Sea Surface Temperature dataset (HadISST1; Rayner et al. 2003). A simple 50-m slab mixed layer model represents all other ocean points. The transition between the prescribed and modeled regions is accomplished through a “blending zone” in which SST is represented by a weighted fraction of prescribed and modeled SST. These integrations are referred to in the text as the pacemaker runs.
To ensure that the ocean remains close to the observed climatology, we prescribe an implied ocean heat-flux or “q flux” field, which is calculated from a 20-yr run using prescribed climatological SST in the pacemaker region and a 60 (W m−2) K−1 relaxation toward climatology in the mixed layer region. The annual cycle of monthly-mean restoring tendencies from the relaxation term is calculated for each gridpoint and introduced as an additional term in the pacemaker mixed layer temperature tendency equation, along with a much weaker [10 (W m−2) K−1] relaxation to climatology. We find that this allows for a relatively small rms error in SST (Fig. 1b) without strongly damping the pacemaker SST variability to climatology. Note that, as we are interested in the ability of the model to reproduce the full global SST pattern, the rms error is calculated for the entire domain, including the pacemaker region. Unless otherwise noted, all model results presented are mean values from an eight-member ensemble, with each ensemble member starting from slightly different initial conditions and covering the period 1950–2002.
The cholera data used in this study consist of monthly symptomatic cases obtained from a surveillance program in Matlab, Bangladesh, conducted by the International Centre for Diarrheal Disease Research, Bangladesh (ICDDR, B). The surveillance site is located 40 km southeast of the capital of Dhaka, in the delta of the Brahmaputra and Ganges Rivers. As noted in the introduction, we focus here exclusively on cholera cases due to the El Tor biotype, as the different strains of cholera exhibit different annual cycles and may be influenced by different aspects of climate variability.
3. Results
The largest values of the correlation occur with SST in the previous winter [December–February (DJF)] and significant correlations are generally confined to the tropical Pacific (Fig. 2). The strongest positive correlations tend toward the central, rather than the eastern, tropical Pacific, most noticeably in December (Fig. 2a) and February (Fig. 2c). We also find significant features in the Indian Ocean and South Atlantic, particularly during January (Fig. 2b). The largest magnitude correlations are negative and occur in the western Pacific off the equator in a horseshoe-shaped region, and not in the eastern Pacific (although these differences are probably not meaningful). This pattern raises the possibility that variations in SST in both the eastern and western Pacific play a role in modulating the regional climate of Bangladesh and potentially cholera risk. However, it is also possible that the pattern in Fig. 2 simply reflects the tendency for negative anomalies in the western Pacific to accompany positive anomalies in the central and eastern tropical Pacific. We address the influence of western Pacific alone on the regional climate of Bangladesh elsewhere (see Cash et al. 2007) and focus here on the influence of the central and eastern tropical Pacific.
There are only minor variations in the correlation pattern between months, indicating that the implied relationship between cholera and SST is not tied closely to a specific lag and is relatively robust. The correlation pattern closely resembles the observed pattern of ENSO composite warm − cold SST anomalies (Fig. 3), consistent with previous results (e.g., Koelle et al. 2005b). We calculate the composite anomalies by averaging over all ensemble-mean El Niño and La Niña events separately and finding the difference between the resulting patterns. As discussed in the introduction, we consider here only those events from 1976 onward (see Table 1 for the list of events), corresponding to the period after the replacement of the so-called classical strain of cholera by the El Tor strain in the environment (Samadi et al. 1983). Taken together with the correlation patterns (Fig. 2) it appears that winter El Niño events enhance fall cholera outbreaks and winter La Niña events suppress fall cholera outbreaks. However, it is also possible that ENSO plays no role in influencing cholera and the correlations arise because of other, confounding variables or to chance.
Examination of average warm and cold ENSO anomalies separately (not shown) indicates a robust linear response in our model in the fields considered here, consistent with previous results (e.g., DeWeaver and Nigam 2002; Frederiksen and Zheng 2007). It is important to recall, however, that while there is a strong linear component the impact of ENSO on the climate system is not fully linear (e.g., Hoerling et al. 2001). There are also significant variations between observed and modeled ENSO events (discussed below). We refer to the warm-minus-cold anomalies as composite anomalies throughout the text.
The composite ensemble-mean DJF SST anomalies (Fig. 4) have significant amplitude not only in the prescribed region in the Pacific but also in widely separated regions around the globe. There are clear negative composite anomalies in the western Pacific horseshoe region in January (Fig. 4b) and February (Fig. 4c). These composite anomalies are generally weaker than observed, particularly in the western Pacific (generally by a factor of 2–3 when compared to Fig. 3). The fact that the composite SST anomalies have significant amplitude outside the central and eastern tropical Pacific indicates that, to a certain extent, SST anomalies in these regions during ENSO years can be considered a response to the forcing in the tropical Pacific. This is consistent with the results of previous pacemaker-type experiments (e.g., Lau and Nath 2003). The pattern correlation between the observed and modeled SST anomalies exceeds 0.9 for the entire domain between 60°S and 60°N, and remains above 0.7 when the prescribed region is excluded; further indication the model is accurately reproducing the global ENSO SST pattern.
One hypothesis for the September–October peak in cholera incidence in Bangladesh is that the extensive flooding during the preceding monsoon months [June–August (JJA)] leads to an increase in cholera incidence by causing breakdowns in sanitary conditions, as well as concentrating the population onto the remaining areas of dry land (A. Dobson, personal communication). Bangladesh is a low-lying country that occupies the delta of two major rivers (the Gangha and Brahmaputra) and is commonly subject to flooding during the monsoon. The dual impact of poor sanitation and overcrowding enhances the probability of coming into contact with water contaminated the bacteria and hence promotes a rapid increase in cholera cases. Under this hypothesis, variations in summer rainfall (Fig. 5) are among the most likely links between the regional climate of Bangladesh and cholera variability. When we compare composite precipitation anomalies following warm and cold ENSO winters, we find that the model appears to support this hypothesis, with large positive composite precipitation anomalies (>3 mm day−1) over Bangladesh in June (Fig. 5a), July (Fig. 5b), and August (Fig. 5c). The composite precipitation anomaly persists into September but is much reduced in amplitude (not shown). Precipitation over the central Indian subcontinent is also somewhat reduced in each month, consistent with the observed weak/anticorrelation between India and Bangladesh rainfall.
As a check on the significance of the ensemble-mean composite precipitation anomalies, we examine the composite JJA precipitation anomalies for each ensemble member (Fig. 6). Although there is some variation in location and intensity, all but one member produces a positive anomaly in excess of 3 mm day−1 in the region 20°–25°N, 85°–90°E. This is a strong indication that the rainfall anomalies in the model arise as a response to the boundary forcing and are not simply a residual of noisy atmospheric processes, and can be taken as a significance test of the model anomalies. It thus appears that there is a definite signal in JJA precipitation over Bangladesh related to the difference in boundary conditions in the central and eastern tropical Pacific. It is worth noting the clear variations in strength and positioning of the anomalies in the individual ensemble members; however, a full analysis and understanding of these variations lies beyond the scope of the current work. We note here simply that the rainfall anomaly exists in some form in each member of the ensemble, confirming the significance of the anomaly shown in Fig. 5.
A question that naturally arises in any modeling study is how well the model represents the observations. We compare the composite precipitation anomaly from the model (Fig. 5) to observed precipitation anomalies (Fig. 7) from the rain gauge–based Chen et al. (2002) product and find that the model appears to capture the general features of the observed response to a winter El Niño event. While there are certain obvious differences in the structure and intensity of the anomaly, particularly in June (Fig. 7a), there is a clear increase in precipitation over Bangladesh in July (Fig. 7b) and August (Fig. 7c), as well as decreases in precipitation over the mainland of India. The anomalies are of similar magnitude in the region of Bangladesh (∼4 mm day−1) in both the model and observations in these latter months. The large increase in precipitation over the west coast of India in July and August is poorly represented in the model (cf. Figs. 5b,c to Figs. 7b,c), which is to be expected given the poor representation of the Western Ghats at T62 resolution. The model also overestimates the reduction in precipitation across central India. However, as the model does reproduce the general features of the observed precipitation anomalies in our region of interest, it is reasonable to explore more fully the mechanism for these changes in precipitation in the model. It should also be noted that there are disagreements among the published research-quality rainfall products in this region (Cash et al. 2008) and the “ground truth” rainfall anomalies may differ somewhat from those shown in Fig. 7. We have chosen the Chen et al. (2002) rainfall product because it includes rain gauge observations within Bangladesh itself and agrees well with the densely observed rainfall values over India (Rajeevan et al. 2005).
Although the model produces a rainfall anomaly that is similar to the observations and consistent between ensemble members an important note of caution is in order. Rainfall does not increase over Bangladesh in every member of the ensemble following every warm event (not shown). Similarly, rainfall does not increase over Bangladesh following every warm event in the observations. As a measure of this variability between events and ensemble members we compare precipitation anomalies in the Bangladesh region for each member of the ensemble for all DJF warm events from 1950–2002. We find that on average 6 of the 8 members will produce anomalies of the same sign and magnitude (not always positive). Thus, while the state of the central and eastern tropical Pacific appears to strongly precondition the system toward a precipitation anomaly of a given sign over Bangladesh, it clearly is not the only factor determining the response.
Variations between ensemble members with the same prescribed SST in the central and eastern tropical Pacific may arise for one of three reasons. They may be entirely due to chaotic atmospheric fluctuations and be thus essentially unpredictable. Alternatively, they may arise due to variations in SST and other boundary conditions outside of the prescribed region. They may also be due to the modulation of the influence of the tropical Pacific by either of the above processes. This significantly complicates the direct comparison of individual events between the pacemaker model and observations, as the model and observations may disagree because of limitations in the model, transient and chaotic atmospheric variability, SST differences outside of the forcing region, or a combination of all three. A full investigation of these potential sources of variability lies outside the scope of the present work and will be treated in detail in future work. In the current work we caution the reader that this variability exists and focus on understanding the processes that give rise to the consistent, forced signal seen in Figs. 5 –7.
One quantity closely related to precipitation is the vertically integrated moisture transport (VIMT) convergence, defined here as the vertical average of −∇ · (uq, υq), where u, υ, and q are the zonal wind, meridional wind, and specific humidity, respectively, from 1000 to 500 hPa (Fig. 8). As expected, there is a close correspondence between the positive and negative precipitation anomalies and convergence and divergence of the moisture transport, respectively (cf. Fig. 8 to Fig. 5), in the region centered on Bangladesh (pattern correlation of 0.6). The precipitation anomalies in Fig. 5 can thus be clearly linked to changes in the monsoon circulation.
When we examine the components of the moisture transport convergence, we find that it is dominated by the convergence of the wind field (see Fig. 9), and that gradients in the moisture field play a limited role (not shown). The vertically integrated zonal wind (Fig. 9) increases upstream of Bangladesh in all three months. The resulting region of convergence lies along the coast of Bangladesh in June (Fig. 9) and then pushes northwest through July (Fig. 9b) and August (Fig. 9c). This progression in zonal wind convergence mirrors the progression in rainfall (Fig. 5) and VIMT convergence (Fig. 8; pattern correlation 0.88), indicating that changes in the zonal wind are playing a key role in the model precipitation anomalies. There is a consistent spatial structure in the zonal wind anomalies in each month, with the strongest values north of 24°N and east of 84°E. We also find a center of divergence in the 200-mb anomalous wind field over the Bangladesh region (not shown), which is likely to be related to the positive rainfall anomalies in Fig. 5.
When we consider the meridional component of the wind field, we see negative wind anomalies throughout JJA immediately upstream of Bangladesh, consistent with a southward deflection of the zonal wind. There is also a clear convergence in the negative meridional winds over the region of enhanced precipitation in June–August (Figs. 5a–c). It is noteworthy that during the months when the positive rainfall anomalies are strongest the meridional wind anomaly is from the land toward the Bay of Bengal. Thus the enhanced precipitation in the model is not due to increased moisture flux off the Bay of Bengal, but rather it is due to the increased moisture flux convergence because of the enhanced westerly zonal winds.
To help us understand how the circulation anomaly in Fig. 9 arises we consider the vertically integrated geopotential height anomaly (Fig. 10) for the monsoon months. In June (Fig. 10a) and July (Fig. 10b) there is a clear gradient in the height field running from northwest to southeast over Bangladesh. Although anomalies in the tropics are weaker in August (Fig. 10c), the remaining anomalies are in the Bangladesh region, consistent with the precipitation anomalies persisting into these months. These circulation anomalies in the Bangladesh region are due a tropics-wide increase in heights following winter El Niño (not shown).
The increase in heights in the tropics during the NH summer months (Fig. 11a) following a winter El Niño represents a continuation through the spring (Fig. 11b) of the wintertime height response to the change in SST (Fig. 11c). During DJF, increases in lower-tropospheric heights are apparent across the tropics with the maximum occurring in the region of Indonesia. This is consistent with a shift in the model’s Walker circulation during El Niño years, with anomalous descent occurring over Indonesia. This is consistent with previous modeling results as well as observations (Alexander et al. 2002) and is further evidence that the pacemaker model is reasonably successful in reproducing the observed ENSO signal. This pattern is weaker in amplitude but still clearly present during the spring (Fig. 11b) and demonstrates that the JJA anomalies (Fig. 11a) are a continuation of the winter anomalies. The summer and winter height anomalies are also reproduced across ensemble members (not shown), consistent with the robust precipitation response (see Fig. 6).
The climate anomalies that give rise to the monsoon season circulation changes (Figs. 4 –11) are in turn a continuation of the winter season temperature anomalies (Fig. 12). There is a strong warming signal in the lower atmosphere that corresponds to the warming at the surface (cf. Figs. 12 and 3). In December (Fig. 12a) the tropical atmospheric anomaly is primarily confined to the eastern tropical Pacific. The anomaly spreads throughout the tropics as the season progresses, and by February (Fig. 12c) the Indian Ocean region is strongly affected.
Taken together, Figs. 4 –12 demonstrate that winter ENSO events give rise to changes in the tropical circulation that lead to changes in monsoon precipitation over Bangladesh, and that the change in model precipitation is similar to what is observed. Warm winter SST anomalies in the eastern and central Pacific lead to a general warming of the tropical atmosphere that persists into summer. This warming leads to a change in the circulation over the Indian Ocean region, which in turn leads to greater moisture convergence and increased precipitation over Bangladesh. These results thus establish a plausible dynamical mechanism by which winter ENSO events can influence fall cholera variability in Bangladesh.
4. Summary and conclusions
In this study we use a regionally coupled, or pacemaker, model to explore the influence of ENSO on the regional climate of Bangladesh. We find that the monsoon rains in the model are enhanced over Bangladesh and suppressed over India following an El Niño event during the NH winter. These changes in rainfall arise from alterations in the overall monsoon circulation, leading to increased convergence over Bangladesh. In the mean, this increase in convergence is due to a weakening of the easterly flow over land near Bangladesh. Following some model El Niño events the anomalous convergence is due to increased northerly flow off the Bay of Bengal. The factors that determine whether an individual El Niño event leads to enhanced northerlies or decreased westerlies is not known at this time and is an ongoing subject of study.
The circulation changes over India and Bangladesh appear to result from a tropics-wide warming of the lower troposphere following an El Niño event. The increased temperatures and the corresponding changes in geopotential heights become prominent over the Indian Ocean in February and slowly decay with the SST anomalies. The height anomalies persist well into the monsoon season and appear to alter the mean monsoon circulation. It is also possible that changes in land surface properties (e.g., temperature, soil moisture, and snow cover) also contribute to the changes in the monsoon.
The link between DJF ENSO events and JJA rainfall in Bangladesh through alteration of the monsoon circulation provides a plausible physical mechanism for the observed relationship between ENSO and cholera shown in Fig. 2 and described in more detail elsewhere. Enhanced rainfall during the monsoon season can lead to increased flooding, facilitating bacterial proliferation and infection. Other effects of the enhanced rains, such as breakdowns in sanitation, may also increase the chances of exposure to contaminated water and hence infection.
The link between the eastern and central tropical Pacific and the climate of Bangladesh indicated by the pacemaker model and the observations raises the possibility that the timing and magnitude of cholera outbreaks in the region can be anticipated. Recent work (Pascual et al. 2008) has demonstrated that including information on the state of ENSO in a susceptible–infected–recovered–susceptible (SIRS) model substantially improves model skill. In addition, improved seasonal forecasts of crop yield and local water availability may also be possible based on this relationship; however, a number of questions remain to be addressed. For example, the internal dynamics of the Indian Ocean may play an important role in influencing the regional climate of Bangladesh and are not accounted for in our model. Likewise, the role of the western Pacific (Cash et al. 2007) and variations between events (B. A. Cash et al. 2007, unpublished manuscript) also require examination in greater depth.
Acknowledgments
Support from NSF Grant EF-0429520 and NOAA Grant NA04OAR4600194 is gratefully acknowledged. XR was also in receipt of funding from the project CIRCE, Climate Change and Impact Research (SUSTDEV-2005-3.L3.1-036961-2) UE, sixth FP. We also wish to express our thanks to Md. Yunus and the ICDDR, B for supplying the data on cholera incidence.
REFERENCES
Alexander, M. A., 1992a: Midlatitude atmosphere–ocean interaction during El Niño. Part I: The North Pacific Ocean. J. Climate, 5 , 944–958.
Alexander, M. A., 1992b: Midlatitude atmosphere–ocean interaction during El Niño. Part II: The Northern Hemisphere atmosphere. J. Climate, 5 , 959–972.
Alexander, M. A., I. Bladé, M. Newman, J. R. Lanzante, N-C. Lau, and J. D. Scott, 2002: The atmospheric bridge: The influence of ENSO teleconnections on air–sea interaction over the global oceans. J. Climate, 15 , 2205–2231.
Bladé, I., 1997: The influence of midlatitude ocean–atmosphere coupling on low-frequency variability of a GCM. Part I: No tropical SST forcing. J. Climate, 10 , 2087–2106.
Bladé, I., 1999: The influence of midlatitude ocean–atmosphere coupling on low-frequency variability of a GCM. Part II: Interannual variability induced by tropical SST forcing. J. Climate, 12 , 21–45.
Cash, B. A., J. L. Kinter III, and X. Rodó, 2007: Links between tropical Pacific SSTs and the regional climate of Bangladesh: Role of the western tropical and central extratropical Pacific. COLA Tech. Rep. 243, 25 pp.
Cash, B. A., X. Rodó, J. L. Kinter III, M. Fennessy, and B. Doty, 2008: Differing estimates of observed Bangladesh summer rainfall. J. Hydrometeor., in press.
Chen, M., P. Xie, and J. E. Janowiak, 2002: Global land precipitation: A 50-yr monthly analysis based on gauge observations. J. Hydrometeor., 3 , 249–266.
Chowdhury, M. R., 2003: The El Niño-Southern Oscillation (ENSO) and seasonal flooding—Bangladesh. Theor. Appl. Climatol., 76 , 105–124.
Colwell, R. R., 1996: Global climate and infectious disease: The cholera paradigm. Science, 274 , 2025–2031.
Cook, G. C., 1996: Management of cholera: The vital role of rehydration. Cholera and the Ecology of Vibrio Cholerae, B. S. Drasar and B. D. Forrest, Eds., Chapman and Hall, 54–94.
DeWeaver, E., and S. Nigam, 2002: Linearity in ENSO’s atmospheric response. J. Climate, 15 , 2446–2461.
Franco, A. A., and Coauthors, 1997: Cholera in Lima, Peru, correlates with prior isolation of Vibrio cholerae from the environment. Amer. J. Epidemiol., 146 , 1067–1075.
Frederiksen, C. S., and X. Zheng, 2007: Variability of seasonal-mean fields arising from intraseasonal variability. Part 3: Application to SH winter and summer circulations. Climate Dyn., 28 , 849–866.
Glass, R. I., S. Becker, M. I. Huq, B. J. Stoll, M. U. Khan, M. H. Merson, J. V. Lee, and R. E. Black, 1982: Endemic cholera in rural Bangladesh. Amer. J. Epidemiol., 116 , 959–970.
Goswami, B. N., 1998: Interannual variations of the Indian summer monsoon in a GCM: External conditions versus internal feedbacks. J. Climate, 11 , 501–522.
Goswami, B. N., V. Krishnamurthy, and H. Annamalai, 1999: A broad-scale circulation index for the interannual variability of the Indian summer monsoon. Quart. J. Roy. Meteor. Soc., 125 , 611–633.
Hoerling, M. P., A. Kumar, and T. Xu, 2001: Robustness of the nonlinear climate response to ENSO’s extreme phases. J. Climate, 14 , 1277–1293.
Ju, J., and J. Slingo, 1995: The Asian summer monsoon and ENSO. Quart. J. Roy. Meteor. Soc., 121 , 1133–1168.
Kirtman, B. P., 2003: The COLA anomaly coupled model: Ensemble ENSO prediction. Mon. Wea. Rev., 131 , 1131–1149.
Kirtman, B. P., and J. Shukla, 2002: Interactive coupled ensemble: A new coupling strategy for CGCMs. Geophys. Res. Lett., 29 .1367, doi:10.1029/2002GL014834.
Kirtman, B. P., Y. Fan, and E. K. Schneider, 2002: The COLA global coupled and anomaly coupled ocean–atmosphere GCM. J. Climate, 15 , 2301–2320.
Klein, S. A., B. J. Soden, and N-C. Lau, 1999: Remote sea surface variations during ENSO: Evidence for a tropical atmospheric bridge. J. Climate, 12 , 917–932.
Koelle, K., and M. Pascual, 2004: Disentangling extrinsic from intrinsic factors in disease dynamics: A nonlinear time series approach with an application to cholera. Amer. Nat., 163 , 901–913.
Koelle, K., M. Pascual, and M. Yunus, 2005a: Pathogen adaptation to seasonal forcing and climate change. Proc. Roy. Soc. London Ser. B, 272 , 971–977.
Koelle, K., X. Rodó, M. Pascual, M. Yunus, and G. Mostafa, 2005b: Refractory periods and climate forcing in cholera dynamics. Nature, 436 , 696–700.
Kohn, G. C., 1995: Encyclopedia of Plague and Pestilence. Wordsworth Reference, 408 pp.
Kraft, C. H., and C. van Eden, 1968: A Nonparametric Introduction to Statistics. The Macmillan Company, 342 pp.
Krishnamurthy, V., and B. N. Goswami, 2000: Indian monsoon–ENSO relationship on interdecadal timescale. J. Climate, 13 , 579–595.
Krishnamurthy, V., and J. Shukla, 2000: Intraseasonal and interannual variability of rainfall over India. J. Climate, 13 , 4366–4377.
Lau, N-C., and M. J. Nath, 2000: Impact of ENSO on the variability of the Asian–Australian monsoons as simulated in GCM experiments. J. Climate, 13 , 4287–4309.
Lau, N-C., and M. J. Nath, 2003: Atmosphere–ocean variations in the Indo-Pacific sector during ENSO episodes. J. Climate, 16 , 3–20.
Misra, V., and Coauthors, 2006: Validating ENSO simulation in coupled climate models. COLA Tech. Rep. 210, 46 pp.
Pascual, M., X. Rodó, S. P. Ellner, R. Colwell, and M. J. Bouma, 2000: Cholera dynamics and El Niño-Southern Oscillation. Science, 8 , 1766–1769.
Pascual, M., M. Bouma, and A. P. Dobson, 2002: Cholera and climate: Revisiting the quantitative evidence. Microb. Infect., 4 , 237–245.
Pascual, M., L. F. Chaves, B. A. Cash, X. Rodó, and M. Yunus, 2008: Predictability of endemic cholera: The role of climate variability and disease dynamics. Climate Res., 36 , 131–140.
Rajeevan, M., J. Bhate, J. D. Kale, and B. Lai, 2005: Development of a High Resolution Daily Gridded Rainfall Data for the Indian Region. Meteor. Monogr. Climatology. No. 22/2005, National Climate Centre, India Meteorological Department, 27 pp.
Rasmusson, E. M., and T. H. Carpenter, 1983: The relationship between eastern equatorial Pacific sea surface temperatures and rainfall over India and Sri Lanka. Mon. Wea. Rev., 111 , 517–528.
Rayner, N. A., D. E. Parker, E. B. Horton, C. K. Folland, L. V. Alexander, D. P. Rowell, E. C. Kent, and A. Kaplan, 2003: Global analyses of sea surface temperature, sea ice, and night marine air temperature since the late nineteenth century. J. Geophys. Res., 108 .4407, doi:10.1029/2002JD002670.
Rodó, X., M. Pascual, G. Fuchs, and A. S. G. Faruque, 2002: ENSO and cholera: A nonstationary link related to climate change? Proc. Natl. Acad. Sci. USA, 99 , 12901–12906.
Samadi, A. R., M. K. Chodhury, M. I. Huq, and M. U. Khan, 1983: Seasonality of classical and El Tor cholera in Dhaka, Bangladesh 17 year trends. Trans. Roy. Soc. Trop. Med. Hyg., 77 , 853–856.
Schneider, E. K., 2002: Understanding differences between the equatorial Pacific as simulated by two coupled GCMs. J. Climate, 15 , 449–469.
Shinoda, T., M. A. Alexander, and H. H. Hendon, 2004: Remote response of the Indian Ocean to interannual SST variations in the tropical Pacific. J. Climate, 17 , 362–372.
Walker, G. T., 1923: Correlation in seasonal variations of weather VIII: A preliminary study of world weather. Mem. Indian Meteor. Dep., 24 , 75–131.
Walker, G. T., 1924: Correlation in seasonal variations of weather IX: A further study of world weather. Mem. Indian Meteor. Dep., 24 , 275–332.
Wu, R., and B. P. Kirtman, 2004a: Impacts of the Indian Ocean on the Indian Summer Monsoon–ENSO relationship. J. Climate, 17 , 3037–3054.
Wu, R., and B. P. Kirtman, 2004b: Understanding the impacts of the Indian Ocean on ENSO variability in a coupled GCM. J. Climate, 17 , 4019–4031.
Wu, R., B. P. Kirtman, and K. Pegion, 2006: Local air–sea relationship in observations and model simulations. J. Climate, 19 , 4914–4932.
(a) Prescribed and blending regions used in pacemaker experiments. Shading denotes weighting of prescribed, observed SST from the HadISST dataset. Weighting is set to 1 in the tropical central and eastern Pacific and the polar region. (b) RMS error (°C) in SST field. Heavy, dark line denotes 12-month running mean.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Lagged rank correlation between September El Tor cholera cases and preceding (a) December, (b) January, and(c) February SST. Shading and contour interval is 0.1. Heavy black contour denotes regions significant at the 90% level.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Observed SST difference for warm − cold ENSO events for (a) December, (b) January, and (c) February. Warmand cold events are based on Climate Prediction Center (CPC) DJF Niño-3.4 index. Shading and contour interval is 0.2°C.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble-mean SST difference for warm − cold ENSO events for (a) December, (b) January, and (c) February. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour interval is 0.2°C.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble-mean precipitation difference for warm − cold ENSO events for (a) June, (b) July, and (c) August. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour intervals are 0.5 mm day−1.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
(a)–(h) JJA mean precipitation difference for warm − cold ENSO events for ensemble members A–H. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour intervals are 1 mm day−1.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
As in Fig. 5, except data are observed precipitation data from Chen et al. (2002) and shading and contour interval is 1 mm day−1.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble-mean vertically integrated moisture transport convergence (1000–500 hPa) difference for warm − cold ENSO events for (a) June, (b) July, and (c) August. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour intervals are 10−9 kg m−1 s−1.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble-mean vertically averaged (1000–500 hPa) streamline difference for warm − cold ENSO events for (a) June, (b) July, and (c) August. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading denotes magnitude.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble-mean vertically averaged (1000–500 hPa) geopotential height difference for warm − cold ENSO events for (a) June, (b) July, and (c) August. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour interval is 1 m.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble-mean vertically averaged (1000–500 hPa) geopotential height difference for warm − cold ENSO events for (a)JJA, (b) MAM, and (c) DJF. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour interval is 2 m.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
Model ensemble mean vertically averaged (1000–500 hPa) temperature difference for warm − cold ENSO events for (a) December, (b) January, and (c) February. Warm and cold events are based on CPC DJF Niño-3.4 index. Shading and contour interval is 0.2°C.
Citation: Journal of Climate 21, 18; 10.1175/2007JCLI2001.1
El Niño and La Niña years used in this study. Events are based on DJF values of the Niño-3.4 region index and are taken from the CPC listing (available online at http://www.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/ensoyears.shtml). Years listed below are for January and February; event definitions include December values from the previous year.
