The influence of meteorological conditions on the extreme 2014 Legionella episode

18:00 Tuesday 28 May




Ana Russo (Portugal) 1; Célia M. Gouveia (Portugal) 2; Pedro M.M. Soares (Portugal) 1; Rita M. Cardoso (Portugal) 1; Ricardo M. Trigo (Portugal)

1 - Instituto Dom Luiz; 2 - IPMA

The transmission of the Legionella disease through the atmosphere is unusual,but not unparalleled. An exceptional outbreak of Legionella occurred in November 2014 in a suburban industrial setting located just north of Lisbon, Portugal, where 375 individuals became ill and 12 died infected by inhalation of steam droplets of contaminated water (aerosols). A standard methodology for studying the propagation of legionella episodes is based on aerosol dispersion models. However, such approaches often require detailed 3-D high resolution wind data over the region, which is not usually available for the needed periods.

The Allwine and Whiteman (AW) approach constitutes an alternative simpler method to assess the assimilative and dispersal capacities of different airsheds [1,4], as it only requires hourly wind components. Thus, it has the advantage of not requiring surface and upper air meteorological observations neither a previous knowledge of the atmospheric transport and dispersion conditions. The objective of this study is to analyze if the legionella outbreak event, which took place in November 2014, had extreme potential recirculation and/or stagnation characteristics. To accomplish the proposed objective, the AW approach was applied to a hindcast time-series covering the affected area (1989-2007) and then for an independent period covering the 2014 event (7-25 November 2014) [5]. Hourly zonal and meridional wind components were retrieved from a regional climate simulation performed using the WRF model over the Iberian domain with a spatial resolution of 9 km. The preliminary results regarding the average daily critical transport indices for the 1989-2007 period clearly indicate that the airshed is prone to ventilation as these events have a dominant presence through most of the study period (72%) %), while only 10% occur due to recirculation and <2% to stagnation. [5]. Comparatively to the 1989-2007 period, the 2014 episode is truly exceptional [5].

[1] Allwine KJ and Whiteman CD., 1994. Atmospheric Environment 28: 713-721.
[2] Blystad H et al., 2005. Eurosurveillance weekly release, 10(5)
[3] Nguyen T et al., 2006. The Journal of Infectious Diseases, 193(1) 102-111.
[4] Mohan M and Bhati S., 2012. Journal of Civil & Environmental Engineering, S1:003.
[5] Russo A et al. 2018. International Journal of Biometeorology, 62: 1167-1179.


This work was partially supported by Fundação para a Ciência e a Tecnologia (Portugal) by the project UID/GEO/50019/2013 – Instituto Dom Luiz. A.Russo thank FCT for the grant SFRH/BPD/99757/2014.