Recent years have seen an increase in the frequency of extreme rainfall and subsequent flooding across the world. Climate change models predict that such flooding will become more common, triggering sewer overflows, potentially with increased risks to human health. In August 2010, a triathlon sports competition was held in Copenhagen, Denmark, shortly after an extreme rainfall. The authors took advantage of this event to investigate disease risks in two comparable cohorts of physically fit, long distance swimmers competing in the sea next to a large urban area. An established model of bacterial concentration in the water was used to examine the level of pollution in a spatio-temporal manner. Symptoms and exposures among athletes were examined with a questionnaire using a retrospective cohort design and the questionnaire investigation was repeated after a triathlon competition held in non-polluted seawater in 2011. Diagnostic information was collected from microbiological laboratories. The results showed that the 3.8 kilometer open water swimming competition coincided with the peak of post-flooding bacterial contamination in 2010, with average concentrations of 1.5x104 E. coli per 100 ml water. The attack rate of disease among 838 swimmers in 2010 was 42% compared to 8% among 931 swimmers in the 2011 competition (relative risk (RR) 5.0; 95% CI: 4.0-6.39). In 2010, illness was associated with having unintentionally swallowed contaminated water (RR 2.5; 95% CI: 1.8-3.4); and the risk increased with the number of mouthfuls of water swallowed. Confirmed aetiologies of infection included Campylobacter, Giardia lamblia and diarrhoeagenic E. coli. The study demonstrated a considerable risk of illness from water intake when swimming in contaminated seawater in 2010, and a small but measureable risk from non-polluted water in 2011. This suggests a significant risk of disease in people ingesting small amounts of flood water following extreme rainfall in urban areas.

The phenomenon of urban flooding due to rainfall exceeding the design capacity of drainage systems is a global problem and can have significant economic and social consequences. This is even more extreme in developing countries, where poor sanitation still causes a high infectious disease burden and mortality, especially during floods. At present, there are no software tools capable of combining hydrodynamic modelling and health risk analyses, and the links between urban flooding and the health risk for the population due to direct contact with the flood water are poorly understood. The present paper outlines a novel methodology for linking dynamic urban flood modelling with quantitative microbial risk assessment (QMRA). This provides a unique possibility for understanding the interaction between urban flooding and health risk caused by direct human contact with the flood water and hence gives an option for reducing the burden of disease in the population by use of intelligent urban flood risk management. The model linking urban flooding and health risk is applied to Dhaka City in Bangladesh, where waterborne diseases including cholera are endemic. The application to Dhaka City is supported by measurements of pathogens in the urban drainage system. The outcome of the application indicates that direct contact with polluted flood water is a plausible route of primary transmission of cholera and demonstrates the applicability and the potential for linking urban flood models with QMRA in order to identify interventions to reduce the burden of disease on the population in Dhaka City.

[1] We present sea surface and upper thermocline temperature records (60-100 year temporal resolution) spanning marine isotope stage 3 ($24-62 ka B.P.) from International Marine Global Change Study core MD01-2378 (121°47.27 0 E and 13°04.95 0 S; 1783 m water depth) located in the outflow area of the Indonesian Throughflow within the Timor Sea. Stable isotopes and Mg/Ca of the near-surface-dwelling planktonic foraminifer Globigerinoides ruber (white) and the upper thermocline -dwelling Pulleniatina obliquiloculata reveal rapid changes in the thermal structure of the upper ocean during Heinrich events. Thermocline warming and increased d18 O seawater (P. obliquiloculata record) during Heinrich events 3, 4, and 5 reflect weakening of the relatively cool and fresh thermocline flow and reduced export of less saline water from the North Pacific and Indonesian Seas to the tropical Indian Ocean. Three main factors influenced Indonesian Throughflow variability during marine isotope stage 3: (1) global slowdown in thermohaline circulation during Heinrich events triggered by Northern Hemisphere cooling, (2) increased freshwater export from the Java Sea into the Indonesian Throughflow controlled by rising sea level from $60 to 47 ka, and (3) insolation-related changes in the Australasian monsoon with associated migration of hydrological fronts between Indian Ocean-and Indonesian Throughflow -derived water masses at $46 -40 ka.

No abstract

Quantitative microbial risk assessments (QMRAs) often lack data on water quality leading to great uncertainty in the QMRA because of the many assumptions. The quantity of waste water contamination was estimated and included in a QMRA on an extreme rain event leading to combined sewer overflow (CSO) to bathing water where an ironman competition later took place. Two dynamic models, (1) a drainage model and (2) a 3D hydrodynamic model, estimated the dilution of waste water from source to recipient. The drainage model estimated that 2.6% of waste water was left in the system before CSO and the hydrodynamic model estimated that 4.8% of the recipient bathing water came from the CSO, so on average there was 0.13% of waste water in the bathing water during the ironman competition. The total estimated incidence rate from a conservative estimate of the pathogenic load of five reference pathogens was 42%, comparable to 55% in an epidemiological study of the case. The combination of applying dynamic models and exposure data led to an improved QMRA that included an estimate of the dilution factor. This approach has not been described previously.