Evaluation of ERA5 precipitation and 10‐m wind speed associated with extratropical cyclones using station data over North America

• Published in: International journal of climatology Vol. 44; no. 3; pp. 729 - 747
• Main Authors: Chen, Ting‐Chen, Collet, François, Di Luca, Alejandro
• Format: Journal Article
• Language: English
• Published: Chichester, UK John Wiley & Sons, Ltd 15-03-2024; Wiley Subscription Services, Inc; Wiley
• Subjects: Bias; Climate studies; Climatology; Cyclones; Earth Sciences; ERA5 evaluation; extratropical cyclone; Extratropical cyclones; Extreme values; local extremes; near‐surface winds; Precipitation; Sciences of the Universe; Wind; Wind speed; Winds; North America
• ISSN: 0899-8418; 1097-0088
• DOI: 10.1002/joc.8339

While the ERA5 reanalysis is commonly utilized in climate studies on extratropical cyclones (ETCs), only a few studies have quantified its ability in the representation of ETCs over land. To address this gap, this study evaluates ERA5's skill in representing the ETC‐associated 10‐m wind speed and the precipitation in central and eastern North America during 2005–2019. Hourly data collected from ~3000 stations, amounting to around 420 million reports stored in the Integrated Surface Database, is used as reference. For the spatial‐averaged ETC properties, ERA5 shows a good skill for wind speed with normalized mean bias (NMB) of −0.7% and normalized root‐mean‐square error (NRMSE) of 14.3%, despite a tendency to overestimate low winds and underestimate high winds. The ERA5 skill is worse for precipitation than for wind speed with NMB of −10.4% and NRMSE of 56.5% and a strong tendency to underestimate high values. For both variables, the best and worst performance is found in DJF and JJA, respectively. Negative biases are often identified over regions with stronger precipitation/wind speeds, and a systematic underestimation of wind speed is found over the Rockies with complex topography. Compared to the averaged ETCs, ERA5's performance deteriorates for the top 5% extreme ETCs with a stronger tendency to underestimate both wind speed and precipitation (NMB of −10.2% and −22.6%, respectively). Furthermore, ERA5's skill is worse for local extreme values within ETCs than for spatial averages. Our results highlight some important limitations of the ERA5 reanalysis products for studies looking at the possible impacts of ETCs.
This study evaluates the ability of ERA5 to represent 10‐m wind speeds and precipitation associated with extratropical cyclones using station observations in North America. For cyclone‐averaged quantities, ERA5 shows a good skill for wind speed while errors are four times larger for precipitation. ERA5 shows strong limitations to represent local extremes within cyclones for both variables. Overall, ERA5 performs the best and worst in DJF and JJA, respectively.