Table 1. Hindcasts for the indicated years and forecast for 2016, for the post-monsoon (Aug-Dec) season of cholera.
The distribution of observed cases for this same post-monsoon period for the training data used to fit the models was used to estimate the values of the 50th (the median), 75th and 95th quantiles. These values are used as thresholds to define outbreaks of increasing size: a season that exceeds the median is considered anomalous, one that exceeds the 75% level, is considered a large outbreak, and one that exceeds the 95% level, an extreme outbreak. The average observed cases for each year are shown next with an indication of whether they exceed the threshold (yes, “outbreak”) or not (no, “no outbreak”). The proportion of 1000 simulations that fall above each threshold level is reported as a probability, and a probability > 50% is interpreted as a prediction of an outbreak, specified in the last column.
| Year | Quantile | Mechanistic temporal model | Statistical spatio-temporal model | ||||
|---|---|---|---|---|---|---|---|
| Observed | Probability | Prediction (prob. > 50%) | Observed | Probability | Prediction (prob. > 50%) | ||
| 2011 | 50 | no | 14.3 | no | no | 2.7 | no |
| 2011 | 75 | no | 1.3 | no | no | 0.5 | no |
| 2011 | 95 | no | 0.1 | no | no | 0.0 | no |
| 2012 | 50 | no | 17.6 | no | no | 17.7 | no |
| 2012 | 75 | no | 2.0 | no | no | 5.1 | no |
| 2012 | 95 | no | 0.1 | no | no | 0.0 | no |
| 2013 | 50 | no | 19.0 | no | no | 43.1 | no |
| 2013 | 75 | no | 1.5 | no | no | 21.2 | no |
| 2013 | 95 | no | 0.0 | no | no | 0.1 | no |
| 2014 | 50 | no | 18.7 | no | no | 33.8 | no |
| 2014 | 75 | no | 1.6 | no | no | 14.2 | no |
| 2014 | 95 | no | 0.0 | no | no | 0.0 | no |
| 2015 | 50 | no | 18.3 | no | no | 56.6 | yes |
| 2015 | 75 | no | 1.8 | no | no | 32.0 | no |
| 2015 | 95 | no | 0.1 | no | no | 0.1 | no |
| 2016 | 50 | -- | 99.0 | yes | -- | 99.6 | yes |
| 2016 | 75 | -- | 86.5 | yes | -- | 97.6 | yes |
| 2016 | 95 | -- | 50.1 | yes | -- | 27.2 | no |