Individual data points don't isolate temporally local events sufficiently. A better measure would be to chart each season as an error-bar range on a graph and

Individual data points don't isolate temporally local events sufficiently. A better measure would be to chart each season as an error-bar range on a graph and visually compare the averages over time to observe trends. I think you might be able to compare that with data at other observation points (airports are usually a good candidate) to show how things likely looked when they weren't under observation.