2010 Annual Report
1a.Objectives (from AD-416)
The objective of this cooperative research project is to conduct geo-spatial analysis of the primary climate and habitat factors that will influence the establishment of Rift Valley fever in the continental United States. A preliminary GIS model already exists; and we will expand this model to the 14 states deemed to be at the highest levels of RVF risk through pathways analysis, and create an online database where users can assess risk in areas and over time periods of specific interest.
1b.Approach (from AD-416)
We will update the model with new laboratory data (provided by outside collaborators) on the temperature-based transmission risk of possible U.S. vectors. If we can establish an accurate and robust enough data source, we will attempt to calibrate the temperature-based model with field data from Kenya. We will add at least one new layer to the risk assessment representing vector habitat. The combination of climate data, habitat data, and livestock density will enable predictions of maximum risk of RVF transmission and establishment in the US. We will create a database of transmission risk data that can be queried for risk analyses at specific locations or times.
Five significant efforts were expended on this project:
1. Historical degree-day data for the 48-contiguous states has been downloaded from DAYMET (a source of small-scale, spatially-interpolated, historical temperatures) and compiled into a database for ease of use with risk mapping. Pictured in Figure 1 is our risk assessment using “best guess” degree-day threshold numbers. These parameters resulted from calibration efforts in CA using WNV bird deaths, and are expected to change pending lab results currently being, which will provide a more accurate number.
2. Livestock data from the 2007 USDA agricultural census has been compiled, linked to a geodatabase, and mapped.
3. Using the two above products we created a map of normalized Rift Valley fever risk taking into account both the number of risk days livestock density. The original scope of work laid out in the proposal called for us to do this work on only selected states, but the program efficiency has been improved and we were able to complete this task for the entire lower 48 states.
a. This product gives the risk of transmission for each day of the year, as well as the annual averages. Data for these efforts were taken from the 1 sq. km. DAYMET database and represent historical norms. We have integrated climate change scenarios into the modeling effort (we have a paper in press and one in review using these techniques) and can readily move further in this direction given different warming scenarios.
4. Degree-day calibration in Kenya using human case data: although these data had good spatial variation, the limited number of meteorological stations in affected areas made this route appear to be a dead end unless we are able to discover some additional sources of historical daily temperature data.
5. Calibration of the degree day model:
a. We are ready to use new lab data as results are finalized.
b. We have made a contact and verbal agreement with a scientist at the New York City Department of Health to use their dead bird data (collected from 2000-2008) in order to calibrate the degree day model using weather stations in NYC. We envision this effort will work in a manner quite similar our past project in Santa Clara County, Ca, and this will be invaluable to see how the model calibrations vary from West Coast to East Coast mosquito populations.
This research supports NP103 Action Plan Components 1. Biodefense Research and 3. Prevent and Control Zoonotic Diseases.
ADODR is directly involved in performance of the research and also monitors activities to evaluate research progress through site visits, meeting at conferences, email and phone calls.