|
Resolution: standard / high Figure 5.
Coherency analysis between cholera incidence, rainfall, and IOI across five countries:
(A) Côte d'Ivoire, (B) Ghana, (C) Togo, (D) Benin, and (E) Nigeria. For each country, incidence data were square-root transformed (SQRT) to rescale
the variance. We obtained mean between selected rainfall time series for three countries:
Ghana, mean between Ghan132 and Ben270, Benin mean between Ben270 and Ben201, and
Nigeria mean between Ben201 and Nige037. To deseasonalize the rainfall time series
all oscillating components with period less than 12 months were removed using a low-pass
Gaussian filter. In addition, all series were normalized. Coherency analyses presented
are, from left to right: incidence SQRT data vs rainfall, incidence SQRT data vs IOI, and rainfall vs IOI. In the coherence power spectra (x-axis: time in year; y-axis: period in year),
power is coded from low value, in dark blue, to high value, in dark red. The white
dashed lines show α = 5% and α = 10% significance levels, computed on 1,000 bootstrapped
series. The inner area, within the cone of influence (black line), indicates the region
not influenced by edge effects.
Constantin de Magny et al. BMC Infectious Diseases 2007 7:20 doi:10.1186/1471-2334-7-20 |