An investigation of relationships between climate and dengue using a water budgeting technique

• Published in: International journal of biometeorology Vol. 45; no. 2; pp. 81 - 89
• Main Author: Schreiber, K V
• Format: Journal Article
• Language: English
• Published: United States Springer Nature B.V 01-07-2001
• Subjects: Animals; Annual variations; Climate; Climate models; Climate variations; Climatic data; Dengue - epidemiology; Dengue - transmission; Dengue fever; Disease Outbreaks - statistics & numerical data; Emergency preparedness; Emergency procedures; Emergency response; Humans; Interannual variations; Modelling; Moisture effects; Prediction models; Predictions; Puerto Rico - epidemiology; Seasonal variations; Seasons; Stochasticity; Vector-borne diseases; Warning systems; Water; Water budget; Water resources; Weather; Puerto Rico
• ISSN: 0020-7128; 1432-1254
• DOI: 10.1007/s004840100090

Climatic water budget indicators were used to assess intra- and interannual variations in dengue incidence for San Juan, Puerto Rico. Multivariate stochastic relationships between dengue and the water budget were developed to (1) determine and quantify the particular climatic measures and their associated lag periods related to dengue variations, and (2) assess their use for dengue prediction and initiation of emergency response procedures. Such a predictive model would be advantageous because it is based on reliable, easily obtained, and low-cost weather observations. Daily running averages of water budget variables over an 8-week period were related to daily running averages of unconfirmed dengue totals (1988-1993). Resultant models show that a variety of water budget and traditional climate measures over the full 8-week period are associated with dengue. The mean seasonal variation in dengue is highly related (R2=88.1%) to the mean seasonal climate variation, with those thermal and energy variables immediately preceding the dengue response showing the strongest relationships. However, moisture variables, predominantly in the form of surplus, are more influential many weeks in advance. For the interannual model (R2=44.1%), energy change, thermal change, and moisture variables are significant across the 8-week period, with moisture variables playing a stronger role than in the intraannual model. Standardisation substantially changes the importance and timing of the variables, and suggests that dengue during this period is more associated with the mean annual variation of climate than deviations from mean conditions. A dengue early-warning model (based on 5 weeks of climate data) was also developed to predict dengue incidence 3 weeks later. While this predictive model is not reliable as a sole predictor of dengue in San Juan, it may be useful as part of a multifaceted watch/warning system.