Subseasonal Prediction of Heat-Related Mortality in Switzerland

Abstract

Heatwaves pose a range of severe impacts on human health, including an increase in premature mortality. The summers of 2018 and 2022 are two examples with record-breaking temperatures leading to thousands of heat-related excess deaths in Europe. Some of the extreme temperatures experienced during these summers were predictable several weeks in advance by subseasonal forecasts. Subseasonal forecasts provide weather predictions from 2 weeks to 2 months ahead, offering advance planning capabilities. Nevertheless, there is only limited assessment of the potential for heat-health warning systems at a regional level on subseasonal timescales. Here we combine methods of climate epidemiology and subseasonal forecasts to retrospectively predict the 2018 and 2022 heat-related mortality for the cantons of Zurich and Geneva in Switzerland. The temperature-mortality association for these cantons is estimated using observed daily temperature and mortality during summers between 1990 and 2017. The temperature-mortality association is subsequently combined with bias-corrected subseasonal forecasts at a spatial resolution of 2-km to predict the daily heat-related mortality counts of 2018 and 2022. The mortality predictions are compared against the daily heat-related mortality estimated based on observed temperature during these two summers. Heat-related mortality peaks occurring for a few days can be accurately predicted up to 2 weeks ahead, while longer periods of heat-related mortality lasting a few weeks can be anticipated 3 to even 4 weeks ahead. Our findings demonstrate that subseasonal forecasts are a valuable-but yet untapped-tool for potentially issuing warnings for the excess health burden observed during central European summers.

Keywords: Switzerland; extreme events; heatwaves; heat‐health warnings; mortality prediction; subseasonal forecast.

Figure 1

Heat‐mortality relationship estimated as best linear unbiased predictions (BLUPs; Methods) and reported as relative risk (with 95$\%$ confidence interval, shaded red) for a cumulative 10‐day lag of summer temperature versus the optimum temperature MMT (temperature of minimum mortality). The results are plotted along the observed temperature range above MMT. The vertical gray dashed and solid lines indicate the MMT and 95th percentile of location‐specific summer temperature, respectively.

Figure 2

Observed daily mean temperature (black solid curve) and daily mean temperature forecasts for the summers of 2018 (left column) and 2022 (right column). The climatological mean and 95th percentile (1990–2017) are given by the black dashed and dotted curves, respectively. The results are shown for forecast initializations during lead week 1 (red), lead week 2 (light green), lead week 3 (blue), lead week 4 (orange), lead week 5 (dark green), and lead week 6 (purple) for the canton of Zurich. The shaded area around the solid line denoting the forecast ensemble mean for a specific lead week indicates the 5th to 95th percentile of the ensemble forecast. The gray vertical shading indicates periods where the forecast‐to‐climatological spread ratio (FCSR) is lower than 0.8, indicating a low intrinsic forecast uncertainty.

Figure 3

Same as Figure 2, but for the canton of Geneva.

Figure 4

Observed (solid black line), climatological (dashed black line), and predicted (solid colored lines) heat‐related mortality fraction in the summer of 2018 (two upper panels) and in the summer of 2022 (two lower panels) for the canton of Zurich. The results are shown for the forecast initializations as described in Figure 2. The shaded area around each lead week's mortality forecast corresponds to the 95$\%$ confidence interval.

Figure 5

Same as Figure 4, but for the canton of Geneva.

Figure 6

Mean error (ME) of predicted heat‐related mortality fraction (i.e., predicted heat‐related mortality fraction minus observed heat‐related mortality fraction) during days with observed temperature above the 95th percentile. The results are given for the cantons of (a) Zurich, and (b) Geneva. The error bars correspond to the 95% confidence interval.

Figure 7

Heat‐related mortality fraction in the cantons of Zurich (top row) and Geneva (bottom row) for the summers of 2018 (left column) and 2022 (right column). The results are given without conditioning on the FCSR (gray bar) and during days with FCSR lower than 1 (brown bar), lower than 0.8 (orange bar), and lower than 0.5 (blue bar). The error bars at each lead week's mortality forecast correspond to the 95$\%$ confidence interval.