CMStatistics 2017: Start Registration
View Submission - CMStatistics
B0266
Title: Linking climate and incidence of dengue in the Philippines with a dynamic spatio-temporal model Authors:  Stephen Jun Villejo - University of the Philippines (Philippines) [presenting]
Rutcher Lacaza - University of the Philippines (Philippines)
Abstract: The effect of climate indicators on epidemics can be confounded with site-specific factors. Modeling of complex systems like the spread of epidemics requires a more adaptive and flexible structure like a dynamic spatio-temporal model proposed. The proposed model is superior over ordinary regression with auto-correlated errors. The model provided additional insights on the distribution of the residuals for space-time points with significant random effects. The tail distribution of the significant group has more variability, which also contains the extreme values. This is especially prominent for rainfall, dew point temperature, and vapor pressure which are important indicators for dengue-carrying vector abundance and population dynamics of mosquitoes.