Intercomparison of selected fixed-area areal reduction factor methods

- Four different ARF methods are compared.
- The ARF methods are implemented using the same dataset over a common region.
- The methods result in different ARF estimates.
- The ARF estimates from each method show significant scatter.
- Results also indicate some ARF dependence on the average recurrence intervals.

SummaryThe areal reduction factor (ARF) is a concept used in many hydrologic designs to transform a point precipitation frequency estimate of a given duration and frequency to a corresponding areal estimate. Various methods have been proposed in the literature to calculate ARFs. Proposed ARFs could vary significantly, and it is unclear if discrepancies are primarily due to differences in methodologies, the dissimilar datasets used to calculate ARFs, or if they originate from regional uniqueness.Our goal in this study is to analyze differences among ARFs derived from different types of fixed-area ARF methods, which are suitable for use with precipitation frequency estimates. For this intercomparison, all the ARFs were computed using the same, high-quality rainfall-radar merged dataset for a common geographic region. The selected ARFs methods represent four commonly used approaches: empirical methods, methods that are based on the spatial correlation structure of rainfall, methods that rely on the scaling properties of rainfall in space and time, and methods that utilize extreme value theory. The state of Oklahoma was selected as the study area, as it has a good quality radar data and a dense network of rain gauges. Results indicate significant uncertainties in the ARF estimates, regardless of the method used. Even when calculated from the same dataset and for the same geographic area, the ARF estimates from the selected methods differ. The differences are more pronounced for the shorter durations and larger areas. Results also indicate some ARF dependence on the average recurrence intervals.