Weather and Attendance

Based on anecdotal comments weather plays an impact in museum attendance. 

In an analysis of past museum attendance data, I can see that Walk Up attendance at my Museum is in fact impacted by weather in a nonlinear way. (For my better models between 8% and 18% of the variability is related to temperature,  3% to 18% humidity, 2% to 14% precipitation.)  I'm interested in making a daily forecast of future attendance using similar models over a 1 - 1.5 year period.  General weather predictions are only good for 7 to 10 days.  So, I need a way to create "Synthetic" weather data daily over a 1 - 1.5 year period into the future that is representative of the kind of weather in my area.  I then hope to do a Monte-Carlo simulation in order to do a sensitivity analysis.  This would give me a range of attendance estimates as well as a most likely estimate.  Questions:

1. Who else has seen weather as significant to your attendance?
   1. Have you been able to do any studies to quantify the influence weather has on your attendance?
      1. Can you quantify the influence?
         1. What level of influence are you seeing?
         2. What are the relationships?
      2. How did you do that study?
2. Is anyone using weather as a part of forecasting attendance?
   1. Are you doing this over the short term (7-10 days) or longer-term (> 14 days)
   2. How are you doing these types of forecasts?
   3. Where do you get your location specific synthetic data for your forecast?

--Tom