Partner 43 – University of Oxford (UOXF)

Oxford University’s Department of Zoology (Oxford, UK) hosts the TALA Research group that was founded by David Rogers in 1993 to examine the role of remote sensing in predicting the distribution and abundance of tsetse flies and the incidence and prevalence of tsetse-transmitted diseases. Today TALA, now part of the larger Spatial Ecology and Epidemiology Group (SEEG) within the Department of Zoology, is a recognized leader in the field of low resolution image processing and applications to a wide variety of diseases. TALA developed temporal Fourier processing techniques to capture elements of habitat seasonality in the NOAA AVHRR satellite data time series, and non-linear discriminant analysis to use the Fourier imagery to describe vector and disease distributions and abundance. This unique combination of image processing and application has now been applied to a wide range of diseases, from anthrax to West Nile Virus locally, regionally and globally. TALA has applied and developed a wide range of time series analytical techniques to examine trends in key seasonality variables across continents and to show changes in the epidemiology of a variety of vector-borne diseases through time.

The same temporal Fourier analysis skills have been applied more recently to composited MODIS data from NASA’s newer Terra and Aqua satellites. These data required the development of new algorithms to cope with the timing of the composited data within each calendar year and the resulting imagery was a major input into the activities and data archives of the EU FP6 EDEN project. They were used, for example, as descriptor variables in the risk maps produced for the EDEN partners inter alia by the Low Resolution Remote Sensing team headed by TALA.

TALA enjoys technical challenges within practical applications, from the use of administrative region rather than point data on disease occurrence to the application of risk mapping methods to describe the distribution of poverty in the Horn of Africa. Risk map production is moving towards using geo-spatial and Bayesian approaches to improve map accuracy. Work has been carried out for a variety of agencies; DFID, DEFRA, DoH in the UK; FAO, WHO, ECDC in mainland Europe; and the INTREPID project in the USA.

TALA/SEEG results are published widely in the reviewed literature, and past and present team members have edited and written chapters for two special issues of Advances in Parasitology (Volumes 47 and 62 for 2000 and 2006 respectively) and contributed to a specialist book on Spatial Analysis in Epidemiology.

Links Page: