Premium
Assessing the importance of spatio‐temporal RCM resolution when estimating sub‐daily extreme precipitation under current and future climate conditions
Author(s) -
Sunyer M. A.,
Luchner J.,
Onof C.,
Madsen H.,
ArnbjergNielsen K.
Publication year - 2017
Publication title -
international journal of climatology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.58
H-Index - 166
eISSN - 1097-0088
pISSN - 0899-8418
DOI - 10.1002/joc.4733
Subject(s) - climatology , precipitation , environmental science , climate model , spatial ecology , downscaling , climate change , pluvial , generalized extreme value distribution , extreme value theory , temporal resolution , flood myth , meteorology , geography , geology , statistics , mathematics , physics , ecology , quantum mechanics , biology , oceanography , archaeology
ABSTRACT The increase in extreme precipitation is likely to be one of the most significant impacts of climate change in cities due to increased pluvial flood risk. Hence, reliable information on changes in sub‐daily extreme precipitation is needed for robust adaptation strategies. This study explores extreme precipitation over Denmark generated by the regional climate model ( RCM ) HIRHAM‐ECEARTH at different spatial resolutions (8, 12, 25 and 50 km), three RCM from the RiskChange project at 8 km resolution and three RCMs from ENSEMBLES at 25 km resolution at temporal aggregations from 1 to 48 h. The performance of the RCM simulations in current climate as well as projected changes for 2081–2100 is evaluated for non‐central moments of order 1–3 and for the 2‐ and 10‐year events. The comparison of the RCM simulations and observations shows that the higher spatial resolution simulations (8 and 12 km) are more consistent across all temporal aggregations in the representation of high‐order moments and extreme precipitation. The biases in the spatial pattern of extreme precipitation change across temporal and spatial resolution. The hourly extreme value distributions of the HIRHAM‐ECEARTH simulations are more skewed than the observational dataset, which leads to an overestimation by the higher spatial resolution simulations. Nevertheless, in general, under current conditions RCM simulations at high spatial resolution represent extreme events and high‐order moments better. The changes projected by the RCM simulations depend on the global climate model ( GCM )– RCM combination, spatial resolution and temporal aggregation. The simulations disagree on the magnitude and spatial pattern of the changes. However, there is an agreement on higher changes for lower temporal aggregation and higher spatial resolution. Overall, the results from this study show the influence of the spatial resolution on the precipitation outputs from RCMs . The biases of the RCM simulations increase, and the projected changes decrease for decreasing spatial resolution of the simulations. This points towards the need for high spatial and temporal resolution RCMs to obtain reliable information on changes in sub‐daily extreme precipitation.