I have to perform a study on runway incursions at JFK airport. The hypothesis is that on rainy days there are more runway incursions then on sunny days. The sample is the weather on the 100 days when runway incursions happened. The random sample from the population is the weather on 100 random days when incursions did not happen.

The question is: do I use a two sample T-test? or a one sample T-test since the second sample is from the population.

Thank you for your time.