Domestic Aedes aegypti breeding site surveillance: limitations of remote sensing as a predictive surveillance tool

This project tested aerial photography as a surveillance tool in identifying residential premises at high risk of Aedes aegypti breeding by extending the use of a recently developed, ground-based, rapid assessment technique, the modified Premise Condition Index (PCI2). During 1995, we inspected 360 premises in Townsville, Australia for Ae. aegypti breeding, and PCI2 scores were recorded. The PCI2 values were also estimated from 1:3,000 color and infrared aerial photograph interpretation for the same premises. We found that shade levels can be accurately identified from both color and infrared images, and the PCI2 can be accurately identified from infrared photographs. Yard conditions, however, cannot be accurately identified from either aerial photograph type. The airborne PCI2 did not significantly correlate with breeding measures, and logistic regression further demonstrated that neither aerial photograph type allows the accurate prediction of Ae. aegypti breeding risk. Therefore, the ability of low-level aerial photography to enhance Ae. aegypti breeding site surveillance is at present limited, with ground surveillance remaining our most reliable tool for identifying the probability of Ae. aegypti breeding in the residential environment.