The influence of air masses on human mortality in the contiguous United States

Abstract

Temperature-related mortality is the leading cause of weather-related deaths in the United States. Herein, we explore the effect of air masses (AMs) - a relatively novel and holistic measure of environmental conditions - on human mortality across 61 cities in the United States. Geographic and seasonal differences in the effects of each AM on deseasonalized and detrended anomalous lagged mortality are examined using simple descriptive statistics, one-way analyses of variance, relative risks of excess mortality, and regression-based artificial neural network (ANN) models. Results show that AMs are significantly related to anomalous mortality in most US cities, and in most seasons. Of note, two of the three cool AMs (Cool and Dry-Cool) each show a strong, but delayed mortality response in all seasons, with peak mortality 2 to 4 days after they occur, with the Dry-Cool AM having nearly a 15% increased risk of excess mortality. Humid-Warm (HW) air masses are associated with increases in deaths in all seasons 0 to 1 days after they occur. In most seasons, these near-term mortality increases are offset by reduced mortality for 1-2 weeks afterwards; however, in summer, no such reduction is noted. The Warm and Dry-Warm AMs show slightly longer periods of increased mortality, albeit slightly less intensely as compared with HW, but with a similar lag structure by season. Meanwhile, the most seasonally consistent results are with transitional weather, whereby passing cold fronts are associated with a significant decrease in mortality 1 day after they occur, while warm fronts are associated with significant increases in mortality at that same lag time. Finally, ANN modeling reveals that AM-mortality relationships gleaned from a combined meta-analysis can actually lead to more skillful modeling of these relationships than models trained on some individual cities, especially in the cities where such relationships might be masked due to low average daily mortality.

Keywords: Artificial Neural Networks; Human Mortality; Synoptic Climatology.

Fig. 1

GWTC2 Air masses (modified from Lee 2016)

Fig. 2

Domain-wide combined relative risk (y-axis) of excess mortality in summer, by lag day (x-axis) for each air mass (black line), along with 95% confidence intervals (gray lines)

Fig. 3

Same as Fig. 2, except for Autumn

Fig. 4

Same as Fig. 2, except for Winter

Fig. 5

Same as Fig. 2, except for Spring

Fig. 6

City-by-city relative risk of excess mortality at Lag1 for each air mass in a summer; b autumn; c winter; d spring

Fig. 6

City-by-city relative risk of excess mortality at Lag1 for each air mass in a summer; b autumn; c winter; d spring

Fig. 6

City-by-city relative risk of excess mortality at Lag1 for each air mass in a summer; b autumn; c winter; d spring

Fig. 6

City-by-city relative risk of excess mortality at Lag1 for each air mass in a summer; b autumn; c winter; d spring

Fig. 7

Same as Fig. 6, except for Lag2 in a summer; b autumn; c winter; d spring

Fig. 7

Same as Fig. 6, except for Lag2 in a summer; b autumn; c winter; d spring

Fig. 7

Same as Fig. 6, except for Lag2 in a summer; b autumn; c winter; d spring

Fig. 7

Same as Fig. 6, except for Lag2 in a summer; b autumn; c winter; d spring

Fig. 8

Averaged output of the 61 separate-city ANN models. Colors/units are Zmort (standardized anomalous mortality) differences between each air mass and the seasonal air mass (thus, for the seasonal AM, Zmort = 0 at all points), averaged by day of the year (y-axis) and lag day (x-axis) after the occurrence of each AM. Output is based on training and validation datasets. Higher/positive values (darker reds) indicate increased mortality relative to that of the Seasonal AM, lower/negative (darker blues) indicate decreased mortality relative to that of the Seasonal AM

Fig. 9

Same as Fig. 8, except using the output of the combined-meta ANN model

Fig. 10

The percentage (y-axis) of 61 cities where meta-model results in that city were more skillful than results for the separate-city model for that city. Skillfulness is measured using root mean squared errors (RMSE, blue) between the model output and actual observed Zmort, and Pearson correlation (orange) between the model output and actual observed Zmort, based on both the training and validation portions of the datasets