Pooled frequency analysis for intensity–duration–frequency curve estimation

Abstract

Rainfall intensity–duration–frequency (IDF) curves are used in the design of urban infrastructure. Their estimation is based on rainfall frequency analysis, usually performed on rainfall records from a single gauged station. However, available at‐site record length is often too short to provide accurate estimates for long return periods. In the present study, a general framework for pooled rainfall frequency analysis based on the index‐event model is proposed for IDF estimation at gauged stations. Pooling group formation is defined by the region of influence approach on the basis of the geographical distance similarity measure. Several pooled approaches are defined and evaluated by a procedure through which quantile estimation and uncertainty are assessed. Alternate approaches for the definition of a pooling group are based on different criteria regarding initial pooling group size (and the relationship between size and return period), approaches for assessing pooling group homogeneity, and the use of macroregions in pooling group formation. The proposed framework is applied to identify the preferred approach for pooled rainfall intensity frequency analysis in Canada. Pooled approaches are found to provide more precise estimates than the at‐site approach, especially for long return periods. Pooled parent distribution selection supported the use of the generalized extreme value distribution across the country. Recommendations for pooling group formation include increasing the pooling group size with increases in return period and identifying an appropriate trade‐off between pooling group homogeneity and size for long return periods.

Estimates of gridded relative changes in 24-h extreme rainfall intensities based on pooled frequency analysis

“…In the present study, the approach is applied to a large case study, consisting of most of Canada, where grid-scale 24-h relative changes are estimated for six return periods (2, 5, 10, 25, 50, and 100 years). Significant trends are found on observed annual maximum rainfall series under local or regional approaches in Canada, with a larger presence of increasing trends (e.g., Burn and Taleghani 2013;Shephard et al 2014;Requena et al 2019). This supports the need of updating IDF curves in Canada based on climate model projections, in line with earlier studies performed at the grid scale for the country (e.g., Mladjic et al 2011;Mailhot et al 2012).…”

“…The pooled estimate of the T-year rainfall intensity quantile is obtained following the recommendations provided by Requena et al (2019), as a result of assessing a number of approaches for pooling group formation according to quantile estimation and uncertainty for pooled rainfall intensity frequency analysis at gauged sites. The approach is applied using "gridyears" instead of "station-years", since grid-scale instead of point-scale annual maximum rainfall intensity series are used in the present study.…”

“…The use of a regional approach to estimate rainfall quantiles from either observations or climate model simulations allows obtaining more accurate estimates, especially for longer return periods (e.g., Mladjic et al 2011;Requena et al 2019), as well as smoother spatial patterns when estimating future changes (Li et al 2017). Some form of fixed-region approach is used for analysing future changes in several studies (e.g., Ekström et al 2005;Mailhot et al 2007Mailhot et al , 2012Mladjic et al 2011;Monette et al 2012;DeGaetano et al 2017).…”

“…The pooled approach not only allows for a more accurate estimation of quantiles associated with longer return periods, but also for more flexibility in the formation of pooling groups for quantile estimation, avoiding marked limits between regions. The pooled approach used in this study is that recommended in Requena et al (2019), which is focused on the gauged case; that is, on the estimation of pooled quantiles at a target site where rainfall series are available. Hence, it is in agreement with information from climate models, since precipitation simulation series are available at the target grid.…”

Reply on RC1

“…While numerous applications of regional frequency analysis are reported for station data (e.g. Gaál and Kyselý, 2009;Requena et al, 2019), fewer examples are available for the extension of times series from weather radar. Goudenhoofdt et al (2017) based a regional frequency analysis over Belgium on pooled radar data time-series with a sampling scheme considering radar cells in a radius of 10 km around the target cell for the extension of the precipitation records.…”