People track, understand, and interpret the weather in order to make more appropriate decisions on a day-to-day basis. Over time, society has become better adapted and more resistant to “poor” weather by diversifying activity choice and adjusting clothing and gear. That said, people choosing to engage in activities with substantial outdoor components are still highly vulnerable to weather fluctuations.

In the event of weather fluctuations, zoos, which have large outdoor components in their experiences, are very exposed to weather-induced attendance variation. The existence of a large hurricane or freezing rain event will definitely shut down a zoo for days to weeks, but the day-to-day fluctuations in the weather make up a majority of yearly uncertainty in predicting attendance values. Economic aspects of weather and climate information become glaringly obvious to the administrators of zoos—and outdoor recreation in general. By utilizing weather and climate information to help predict attendance values (demand), businesses can increase their operating margins through smarter organizational decisions. Such abilities of zoos include adequate staffing levels for the amount of daily demand, timing of exhibit closures to impact the fewest attendees, appropriate health-stress monitoring to prevent medical emergencies, and the availability of vending.

This study, which works with daily attendance data at the North Carolina Zoo in Asheboro, North Carolina and Zoo Atlanta in Atlanta, Georgia, analyzes the predictability of attendance using weather and climate data.

The questions that drive this research are as follows:

•How susceptible are zoo attendances to everyday weather events?

•Can zoos increase the accuracy of their attendance prediction models with weather data?

•How do weather influences change between rural (NC Zoo) and urban (Zoo Atlanta) zoos?

Using these research questions as a guideline and working in concert with both the North Carolina Zoo and Zoo Atlanta, predictive attendance models are developed. These models use multiple regression analysis and are tailored seasonally for each zoo using varying temporal scales while still factoring for weekend, weekday, and holiday “social drivers.” Weather variables are separated into four general categories of prediction but included singly as independent variables in the multiple regression models. The four categories used are “common variables” such as temperature and precipitation, “underlying variables” which include enhancing factors such as cloud cover, humidity, and wind speed, “index based variables” which are scientifically derived heat balance outputs, and “relative variables” which assess departures from “typical” weather such as a cool summer day. The resulting models are analyzed for explanation of variance (R-squared) and compared to decipher different responses to varying weather variables based on the social structure and demographic of an “attendee” at a particular zoo.

Social forces are admittedly the strengths of attendance prediction; however, the subtle differences in a zoo's weather vulnerability—whether it is related to city proximity, spatial layout, surrounding demography, or climate of the location—are factors that underlie the scholarship of how weather and weather perception both affect zoo attendance.