Using Probabilistic Rainfall and Probabilistic River Information to Provide Long-Lead IDSS for Record Flooding along the Illinois River Basin in Eastern Oklahoma during the 2015 Christmas Holiday

Widespread totals of eight to twelve inches of rain fell in a two-day period across the Illinois River Basin of eastern Oklahoma and northwest Arkansas, resulting in devastating, record flooding along the Illinois River on December 28, 2015. Using locally determined worst-case rainfall amounts in conjunction with probabilistic river forecast information from the National Weather Service (NWS) Arkansas-Red Basin River Forecast Center (ABRFC), the NWS Weather Forecast Office (WFO) in Tulsa, Oklahoma determined days in advance that the reasonable worst-case scenario was for an unprecedented flood event over the Christmas holiday. It was important to begin coordination efforts with public safety officials about the potential of a record flood event well ahead of the typical time frame, given the urgent message would be diminished for those focused on Christmas activities and the fact this would be an atypical event for this time of year.

WFO Tulsa staff provided aggressive Impact-based Decision Support Services (IDSS) due to the impact potential and need for preparations on and before Christmas Day. WFO Tulsa engaged partner agencies by briefing them an unprecedented five days ahead of the rain and seven days before the peak flooding. WFO Tulsa and the ABRFC have pioneered techniques in probabilistic forecasting for events like this one. The potential for record river levels was beyond the time frame of the official river forecast, and while the deterministic rainfall forecast indicated high rainfall totals, more extreme values seen in the probabilistic-QPF were also thought possible locally. Therefore, using probabilistic data was the best way to effectively communicate the threat. IDSS intensified as the event approached, including live webinars and one-on-one briefings to local officials on Christmas Day, and continued for the duration of the flooding. WFO Tulsa used enhanced life-threatening wording in all messaging, which triggered evacuation of homes and emergency protective actions for property and livestock. Officials attributed the lack of fatalities to particularly accurate messaging from the NWS.

This presentation will discuss the probabilistic tools developed by WFO Tulsa and ABRFC, illustrate how they were used to support WFO Tulsa’s IDSS efforts with partner agencies, and describe how decision makers acted on this information to protect life and property in their jurisdictions.

*Corresponding author address: Nicole M. McGavock, NWS Tulsa 10159 E. 11^th St. Suite 300, Tulsa, OK 74128; email: nicole.mcgavock@noaa.gov