Fig 1.
Time series and spectral analysis plots for NYC turbidity, source water turbidity, and diarrhea ED visits from 2002–2009.
Fig 2.
Season specific power spectra of median NYC turbidity, 2002–2009.
Fig 3.
Box plot of day-of-week patterns for median NYC turbidity and diarrhea ED visits by season, 2002–2009.
Fig 4.
Cross Correlation Function (CCF) results for turbidity and diarrhea ED visits.
CCF results for 2002–2009 for A) NYC turbidity and diarrhea ED visits all ages B) NYC turbidity and diarrhea ED visits 0–4 age group C) NYC turbidity and diarrhea ED visits 5–17 age group and D) Source water turbidity and diarrhea ED visits all ages. Each bar shows the degree of correlation between ambient temperature and diarrhea ED visits. Solid bars to the right of the vertical dashed lines indicate positive correlations, hollow bars to the left indicate negative correlations. The bottom half of each panel (below the center line at lag 0), represents where the first variable leads the second and the area above the center line at lag 0 indicates that the second variable leads the first.
Fig 5.
Regression results for turbidity and diarrhea ED visits.
Regression results for A) NYC turbidity and diarrhea ED visits for all ages and B) Source water turbidity and diarrhea ED visits (2002–2009). The regression analysis between turbidity and diarrhea ED visits for all ages shows 0% excess risk (ER) at day 0 and an increasing %ER which peaks at 4% at a lag of day 6 in the spring season. There is no consistent increased excess risk in other seasons.
Fig 6.
Sensitivity of diarrhea ED risk estimate at lag day 6 to removal of up to the highest 20% of turbidity values, 2002–2009.