In theory, the mosquito population and weather models should be more realistic. How would you improve them?
How would you model infection routes other than mosquitoes biting birds? For example, what if mosquitoes can infect their eggs with the virus? What if birds can get it form eating other birds that died from the disease?
What if the birds that survive are more likely to die in a drought? That is, if recovered birds die, how does the chance that a mosquito bites an infected bird change?
How would you test whether the new models were more accurate?
What are the multi-year effects? For example, the second year many more birds will be immune. A few years with good weather might mean that few birds were infected so fewer are immune; an El Nino weather cycles could increase West Nile Virus.
If the season when mosquitoes are active is longer, how does the model need to change?
One of the main ways to control diseases carried by mosquitoes is to reduce the number of places mosquitoes can lay their eggs. That is, people need to clean out their birdbaths more often, empty saucers under flower pots, and pick up trash. Can you model the effectiveness of these activities?
One of the reasons to make models is to help decide which experiments to do first. For example, should we get better data on how long different kinds of birds stay sick or on how likely a bird is to get sick from one mosquito bite? If small changes in one assumption make a big change in the outcomes, we want to know more about that assumption.