Attributable deaths due to urban heat island effect in a mega city of Vietnam: an application of dynamic downscaling weather model

Abstract

Introduction: Previous studies have examined the intra-city variation in heat-related mortality; of these, however, no studies directly quantifies magnitude of the urban heat island (UHI) risks to mortality. The purpose of this study is to investigate the attributable deaths due to UHI with-in Ho Chi Minh (HCM) city, a mega city of Vietnam using dynamic downscaling weather model. Methods: The analysis consists of the following steps: (1) We used dynamic downscaling method, accounting for urbanization factors (i.e. urban size, urban surface characteristics, anthropogenic heat, topography, etc.) to estimate spatial temperatures of each districts with in HCM city. (2) For each districts we calculated mortality attributable fractions (AFs) due to total heat, extreme-heat, and mild-heat, following the previous method by Gasparrini et al. (3) The subtract of AFs due to total heat between central districts (centers) and outer districts (outers) is then calculated, which we call as AF due to UHI effect. Results: Overall, centers were hotter and drier compared to outers. The mean of average temperature of centers was 0.9oC higher compared to outers (28.4oC vs. 27.5oC); whereas, the means of average relative humidity in centers and outers were 68.6%, and 75.1% respectively. In addition, number of hot days (average temperature >= 30oC) was higher in centers compared to outers (108 days vs. 42 days). The AFs due to total heat, extreme-heat, and mild-heat were 1.42%, 0.3%, and 1.12% respectively in centers; and were 1%, 0.26%, and 0.74% respectively in outers. Therefore, the AF due to UHI effect was 0.42%. Conclusions: The study found a difference in weather conditions, and AFs due to heat components between central districts and outer districts. The AF due to UHI effect in HCM city was 0.42%. This information is valuable for authorities in considering how much the UHI effect on mortality may be minimized by implementing appropriate planning and intervention.