Spatiotemporal dynamics of dengue in Colombia in relation to the combined effects of local climate and ENSO

Abstract

Dengue virus (DENV) is an endemic disease in the hot and humid low-lands of Colombia. We characterize diverse temporal and spatial patterns of monthly series of dengue incidence in diverse regions of Colombia during the period 2007-2017 at different spatial scales, and their association with indices of El Niño/Southern Oscillation (ENSO) at the tropical Pacific and local climatic variables. For estimation purposes, we use linear analysis tools including lagged cross-correlations (Pearson test), cross wavelet analysis (wavelet cross spectrum, and wavelet coherence), as well as a novel nonlinear causality method, PCMCI, that allows identifying common causal drivers and links among high dimensional simultaneous and time-lagged variables. Our results evidence the strong association of DENV cases in Colombia with ENSO indices and with local temperature and rainfall. El Niño (La Niña) phenomenon is related to an increase (decrease) of dengue cases nationally and in most regions and departments, with maximum correlations occurring at shorter time lags in the Pacific and Andes regions, closer to the Pacific Ocean. This association is mainly explained by the ENSO-driven increase in temperature and decrease in rainfall, especially in the Andes and Pacific regions. The influence of ENSO is not stationary (there is a reduction of DENV cases since 2005) and local climate variables vary in space and time, which prevents to extrapolate results from one site to another. The association between DENV and ENSO varies at national and regional scales when data are disaggregated by seasons, being stronger in DJF and weaker in SON. Specific regions (Pacific and Andes) control the overall relationship between dengue dynamics and ENSO at national scale, and the departments of Antioquia and Valle del Cauca determine those of the Andes and Pacific regions, respectively. Cross wavelet analysis indicates that the ENSO-DENV relation in Colombia exhibits a strong coherence in the 12 to 16-months frequency band, which implies the frequency locking between the annual cycle and the interannual (ENSO) timescales. Results of nonlinear causality metrics reveal the complex concomitant effects of ENSO and local climate variables, while offering new insights to develop early warning systems for DENV in Colombia.

Climatic factors and the incidence of dengue in Cartagena, Colombian Caribbean Region

“…Our results suggest that macroclimatic phenomena do not play a role in dengue dynamics in Cartagena. These findings are consistent with those previously reported in the Caribbean region and Bolívar Department, where no significant correlations between dengue cases, macroclimatic phenomena, and climatic variables were found 3 . However, in contrast to the observations at these higher geographic scales, climatic variables and dengue cases were correlated in Cartagena.…”

“…In Colombia, dengue cases were negatively correlated with precipitation and positively correlated with the temperature at the national level. Furthermore, the El Niño phenomenon correlated with increased cases in the country 3 . However, these correlations may vary at lower spatial scales, such as regional, departmental, and municipal scales, and extrapolation of relationships between climate variables or phenomena and the number of dengue cases from one scale to another would be inappropriate without prior evaluation 3 .…”

“…Furthermore, the El Niño phenomenon correlated with increased cases in the country 3 . However, these correlations may vary at lower spatial scales, such as regional, departmental, and municipal scales, and extrapolation of relationships between climate variables or phenomena and the number of dengue cases from one scale to another would be inappropriate without prior evaluation 3 . In the Caribbean region of Colombia, there are several tropical dengue-endemic areas; however, the relationship between dengue incidence and weather factors has hardly been described in the municipalities of this region 4 .…”

“…Overall, a positive relationship between dengue cases, rainy days, and precipitation is expected, as increased rainfall creates more breeding sites for mosquitoes and larval development 7 . However, it is possible that, as in other Colombian areas, warm and dry conditions interact with social-ecological, biological, and entomological factors to play an important role in the incidence of dengue cases in Cartagena 3 .…”

Causal discovery of drivers of surface ozone variability in Antarctica using a deep learning algorithm

“…Further details about the PCMCI can be found in Runge et al 38 Recently, PCMCI + , an extension to PCMCI, has been proposed, which allows the discovery of contemporaneous links in addition to the lagged links up to a Markov equivalence class. 39,40 We examine the stationarity of our datasets using the Kwiatkowski-Phillips-Schmidt-Shin (KPSS) test and Augmented Dickey-Fuller (ADF) unit root test before performing causal discovery using PCMCI and stationaries the dataset by the rst-order differencing as required. We take eight months as T max to account for tropical teleconnections to the polar region, and a is taken as 0.05.…”