Dynamic Continuous Area Spacie-Time (DYCAST) System for West Nile Virus Prospective Analysis
The West Nile virus (WNV) is a mosquito-borne disease-causing infectious agent that affects wildlife and domestic animals. Birds act as an amplifying host and their death from the virus presages the spread of the virus to human populations. Most states set up call in numbers for public reporting dead birds. Bird locations and species are input to a data-based. There data are sent to the DYCAST Model for processing. Results are displayed on an internet site and can be downloaded into a local GIS.
Public: for crowd sourcing location of dead birds (i.e. 1-800-DEAD-BIRDS)
State Health Officials: for compilation and geo-coding of data and transfer to DYCAST model.
Mosquito control boards: that use model for monitoring and remediation.
Calibration tool --> Kappa Analysis --> DYCAST model run --> reporting tool
WNV activity in the previous year that could be used for calibrating the model
Workflow is as follows: Calibrate model from previous years dead-bird data. Parameters will include: (1) size of spatial- temporal domain; & (2) size of spatial-temporal close-in-space-close-in-time sub-domain. Run Monte-Carlo simulation to obtain an empirical distribution for the combination of the spatial-temporal domain and its close-in-space-close-in-time sub-domain. Use a spatial-temporal Kappa analysis to find the lag and window size that represents the period of maximum predictability of human (risk of) infection. Run the model prospectively starting at the beginning of the season. Report DYCAST areas of risk (based on a grid size of 1/4 mile) to mosquito control boards for monitoring and remediation using a web-based portal.
Not sure about this. The model runs every day of the WNV season (May to October). Results from each day can be used in a post-season analysis for validation. For seasons 2005-2008 model was run every day for every 1/4 x 1/4 mile of the State of California.
Other data can enter into the model such as real-time meteorological data, demographic layers, soils, land-use layers and hydrography layers.