While WNV is a disease that can cause severe illness it is fortunate that it does not spread via person to person contact. The primary vector for spreading the disease is by mosquito bites. The city of Chicago takes this threat seriously and for many years has worked to learn about the spread of the disease by gathering data about mosquitos from traps located at many locations across the city.
To best target WNV mosquitos, our team built a classification algorithm that will help the city of Chicago conserve resources while mitigating the threat.
The daily data includes location of the traps and the number of WNV bearing insects in each trap. This data, along with weather data and an understanding of the life-cycle of a mosquito, can help the city know when is the best time and place to spray pesticides to prevent a flare up of WNV in Chicago.
Given the data provided to us, we wanted to most accurate predict when and where the city of Chicago should spray for West Nile, while at the same time being cognizant of limited resources to do so.
There are numerous scientific publications that discuss the tracking of mosquitos. We refer you to “Predicting Culex pipiens/restuans population dynamics by interval lagged weather data” by Karin Lebl Katharina Brugger and Franz Rubel (see https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3660179/) in the journal Parasites & Vectors. This paper influenced our choices of input variables, as it determined that the length of day was a key indicator of West Nile Virus carrying mosquitoes.
Mosquito populations will spike in high temperature and high humidity months.