Parameters of multi-linear regressions for reconstructing peak flow in Contiguous United States (CONUS)

This dataset contains multi-regressions for 672 gauges across Contiguous United States (CONUS) to extend the peak dataset for enhanced Flood Frequency Analysis (FFA).

Flooding is a recurrent natural disaster causing substantial damage and casualties worldwide. A critical task to prevent and mitigate the negative impacts of these natural hazards is to characterize the frequency of flood peaks – a process known as flood frequency analysis (FFA). However, the short records of peak flow observations often limit the FFA accuracy. Here, we developed a statistical method to expand peak flow records at 672 undisturbed gauges across the United States using observations of daily mean flow, available over relatively long periods. We also quantified how FFA reliability improves by adding these expanded datasets of peak flows. This work provides datasets and benchmarks for increasing FFA accuracy, which are helpful for practitioners and government agencies responsible for flood mitigation, infrastructure design, and water management in the United States.