Concurrent Functional Regression to Reconstruct River Stage Data During Flood Events

• Activity Number: 413 - Analyses of Environmental Data
• Type: Contributed
• Date/Time: Thursday, August 12, 2021 : 2:00 PM to 3:50 PM
• Sponsor: Section on Statistics and the Environment
• Author(s): Ryan Pittman*
• Companies: University of South Carolina Department of Statistics
• Keywords: Functional Data Analysis; Functional Prediction Intervals; Gage Height; Hydrology; Landmark Alignment

Abstract

On 10/4/15, the Cedar Creek gage at Congaree National Park stopped reporting stages during some of the worst flooding in SC history. Our goal is to reconstruct the missing stages. The Congaree River gage remained functioning and reached its maximum recorded crest. The stages from the two gages are directly related during floods as water travels through the local spillways to connect them. We introduce a new method called Landmark Aligned L1 distance to determine the span of each of the top ten historical flood events and use these events to reconstruct the missing Cedar Creek stage, greatly improving the reconstruction’s accuracy. We treat the stages as functional data and use a concurrent functional model to establish the relationship between the two locations. Then, the known October 2015 Congaree stage is used to reconstruct the missing Cedar Creek stage for the 2015 flood. The results show that the novel LAL1 distance data selection method is effective, and there is a strong functional relationship between the two locations. We estimate that Cedar Creek reached a historic high in October 2015, with stages exceeding 17 feet, compared to a previous high of just over 16 feet.