Table 3.
Multivariate analysis of time series*
| Arima without climatic variables | (AIC = 397.4) | ||
|---|---|---|---|
| Parameter | Standard deviation | P | |
| Autoregressive term | 0.76950 | 0.07500 | 0.00000 |
| Moving average term | −0.27920 | 0.10730 | 0.00463 |
| Seasonal moving average term | −1.00000 | 0.14380 | 0.00000 |
| Arima with climatic variables | (AIC = 349.57) | ||
| Autoregressive term | 0.57720 | 0.11430 | < 0.00001 |
| Moving average term | −0.11260 | 0.13670 | 0.20498 |
| Seasonal moving average term | −0.99990 | 0.12030 | 0.00000 |
| Cumulated rainfall (2 months lag) | −0.00100 | 0.00030 | 0.00012 |
| Temperature (8 months lag) | 0.22270 | 0.07820 | 0.00221 |
| Number of days with > 50 mm of rain (4 months lag) | −0.06970 | 0.02800 | 0.00649 |
| Number of days with > 50 mm of rain (7 months lag) | −0.10090 | 0.02840 | 0.00019 |
| MEI “El Niño” index (4 months lag) | 0.23910 | 0.08170 | 0.00172 |
*
Parameters of ARIMA models and climatic variables significantly correlated with leishmaniasis. Improvement of model by climatic variables (Akaike's information criterion [AIC] fraction): 1- (349.57/397.4) = 12% Ljung-Box white noise test: P = 0.9511.