The team used real - time
seasonal rainfall, temperature and El Niño forecasts, issued at the start of the year, combined with data from active surveillance studies, in a probabilistic model of dengue epidemics to produce robust dengue risk
estimates for the entire year.