The categorization of wind shift events using metar and personal weather station data

Every day airport managers consistently analyze current and future weather conditions to determine whether their facility will be negatively impacted. From thunderstorms, to fog, to snow and ice, there is a multitude of weather events that can decrease the number of planes that an airport can handle in a given hour. One of the more important weather factors, and often overlooked, is wind, specifically wind shifts. On a daily basis, an airport will set its runway configuration based on the expected dominant wind flow across the area in order to maximize the efficiency of the terminal area. If the wind does not change direction over the course of the day, the airport is able operate at its optimum level, barring any other impactful weather event or a constant “bad” wind direction. If the wind does shift its direction, a change in the airport´s runway configuration is required. This decision of when to change the runway configuration, however, is not always easy, and often times it can be a difficult and sometimes costly one. If the configuration of the runway is changed too late or too early in relation to the time of the wind shift, the throughput at the airport will decrease. To better understand these wind shifts, two different sources of weather observation data were investigated: METAR and Personal Weather Station (PWS). For all of 2012, METAR data for four airports were downloaded from the National Weather Service, and PWS data for those stations within 10 miles of Hartsfield-Jackson Atlanta International Airport (ATL) were downloaded from Weather Underground. The wind direction and wind speed observations were analyzed from both data sources to identify the occurrence of wind shift events at each station, and then each event was categorized based on its length. This study shows the differences in the length of wind shifts between stations, but also illustrates the advantages and disadvantages of using these data sources for categorizing wind s- ift events.