I am a theoretical and computational physicist. For many years, I worked on electronic transport in disordered
materials. Essentially, this entailed using a theoretical/computational approach to problems in quantum mechanics of interacting systems. I
wrote a number of papers on that subject. (See my curriculum vitae , publications 1-22,
presentations 1-25, and all externally funded grants.) In the last few years, however, I have become interested in the following problem,
where an interdisciplinary approach might be beneficial.
Statistical modeling of valley fever incidence. Coccidioidomycosis (valley fever) is a fungal
disease acquired through the inhalation of spores of Coccidioides immitis or
Coccidioides posadasii. The disease afflicts humans and other mamals, and is endemic
to many areas in the
Americas including much of the southwestern United States and northern Mexico. There have
been several attempts at defining the ecological niche of Coccidioides spp.
However, the details of the fungus's environment have escaped full characterization to this date, most probably
because it has been very difficult in the past to isolate and identify Coccidioides spp.
from soil samples (and impossible from air samples).
Importantly, it is known that the fungus does not grow in disturbed (e.g. cultivated) soils.
It is natural to surmise that
environmental conditions which foster the growth and dispersal of Coccidioides spp.
contribute to valley fever incidence. Indeed, in a number of
publications,
we endeavored to
establish a connection between climatic variations on one hand, and valley fever incidence
fluctuations and outbreaks on the other. We obtained some important results and insights, but
our model outputs indicate only a weak connection between incidence and environmental fluctuations.
See figure 1.
We think that this is because the problem is a complex one (in the mathematical sense) with several
interacting factors. Thus, the true nature of the mechanism for valley fever transmission will
only emerge when we incorporate into our models all the important disease incidence drivers.
Figure 1. Valley fever incidence (number of cases per week per 100, 000 population) in Kern County in green for January 1995 to December 2003. Model predictions in red. The main predictor in the model plotted in this graph is incidence at prior times, not any climatic parameter. Since valley fever is not contagious, this indicates that we are missing predictors in our model.
Figure 2. False color image of the southern end of the San Joaquin Valley on May 12, 2010 at 2:52 PM local time. Yellow pixels indicate locations with C. immitis-friendly environments. Upper circle indicates the City of Delano, center-right circle denotes the City of Bakersfield, and center-left circle indicates the City of Taft.
We work in close collaboration with faculty in the Departments of Biology and Mathematics, and with personnel at the Kern County Department of Public Health. Our goals are to investigate the ecology of C. immitis, and to construct products which can be used as public health advisory tools.
Publications related to this subject
Zender, C. S., & Talamantes, J. (2006).
Climate controls on valley fever incidence in Kern County, California,
International Journal of Biometeorology 50, 174 (2006), PDF,
doi: 10.1007/s00484-005-007-6
Talamantes, J., Behseta, S., & Zender, C. S. (2007).
Statistical modeling of valley fever data in Kern County, California, International
Journal of Biometeorology 51, 307 (2007) PDF
doi:10.1007/s00484-006-0065-4
Talamantes, J., Behseta, S., & Zender, C. S. (2007).
Fluctuations in climate and incidence of Coccidioidomycosis in Kern County, California: a review,
Annals of the New York Academy of Sciences 1111, 73 (2007), PDF
doi: 10.1196/annals.1406.028.