•  
  •  
 

Abstract

We developed a two-scale time series framework using monthly container throughput (20-foot equivalent units, TEU) as a port activity indicator. Applied to Kaohsiung, Taiwan (2017–2024), we related TEU to near-port NO₂ and PM₂.₅ (monthly) and air quality to outpatient respiratory diagnoses (weekly) across the 2019–2020 sulfur rules and the COVID-19 pandemic. Transfer-function models indicate TEU tends to lead near-port NO₂ by 6–9 months. Structural break tests at policy dates indicate nonsignificant changes in levels and slopes, suggesting gradual rather than abrupt shifts in the activity–pollution linkage. Weekly two-pollutant distributed-lag Poisson models estimated relative risks (lags 0–8) per interquartile range (IQR) increase. For acute upper respiratory infection, NO₂ was associated with higher risk (RR, 1.184; 95% CI, 1.032–1.358), whereas PM₂.₅ estimates were null (RR, 1.032; 95% CI, 0.917–1.161). Pre-specified negative-control endpoints yielded mixed-sign associations, which were interpreted as residual design noise rather than a coherent nonrespiratory signal. Combining TEU statistics with meteorology-adjusted NO₂ helps screen whether maritime policies and operational changes coincide with air-quality shifts and respiratory risk. This approach requires only publicly reported TEU statistics, routine air-quality monitoring, and syndromic outpatient counts, and is therefore feasible in settings where AIS-based ship-activity metrics or detailed emissions inventories are unavailable.

Abstract Image

Additional Files
JMARSCITECHTAIW-D-25-00078R1_Supplement_R2.pdf (1199 kB)
Supplementary Material
Download
COinS