Abstract
Despite their fame, Ebola and Marburg viruses (family Filoviridae) remain mysterious. Filovirus outbreaks are restricted to tropical Africa, but their likely geographic extent has been outlined only recently, and their natural reservoir host(s) remains unidentified. If environmental conditions associated with outbreaks in space and time can be identified precisely, much could be learned about the ecology, evolution and transmission of these viruses. We examined five filovirus outbreaks for which time series of remotely sensed data (NDVI values) are available and for which the reservoir-to-human index case transmission timing and location are known. A wavelet analysis was used to detect anomalous behaviour in the NDVI signal across multiple time scales for each outbreak. Scale-wise anomalies were manifested at a 20-day scale, and were found in four of the five sites, one to three weeks preceding outbreaks, suggesting that filovirus-caused disease outbreaks may be associated with either behavioural shifts in a vertebrate host or changes in viral population dynamics. More generally, wavelet analysis offers a powerful tool for identifying temporal correlates of disease transmission events.
Acknowledgements
The authors thank David Vieglais, Kevin Ruland and Rob Gales, for their help with manipulating the RS data. Additional thanks to Karl Johnson, Jim Mills and Darin Carroll for their sage advice and input. Data used were generously provided by the Earth Observing System Pathfinder Program of NASA's Mission to Planet Earth, in cooperation with the National Oceanic and Atmospheric Administration. Work was done under a contract from the US Department of Defense.