An assessment of ERA5 total column water vapour in extratropical cyclones relative to the ESA-CCI WV CDR2 product

Flooding and extreme precipitation events in the mid-latitudes, often driven by extra-tropical cyclones (ETCs), are among the most significant meteorological hazards linked to climate change. These events typically arise from the combination of strong surface winds associated with cyclones and high values of atmospheric moisture, in particular total column water vapour (TCWV), which is projected to increase under global warming scenarios. However, only few studies have systematically analysed the moisture environments of ETCs in reanalysis, and direct comparisons with observational datasets, such as satellite products, are scarce. 

This study investigates the potential of the ESA Water Vapour Climate Change Initiative (ESA-CCI WV) CDR2 product to evaluate TCWV associated with ETCs in the ERA5 reanalysis in 2002-2017. Particular emphasis is placed on identifying water vapour biases that affect the simulation of mid-latitude precipitation extremes. Northern Hemisphere ETCs are tracked in the 6 hourly minimum mean sea level pressure, smoothed to T63 resolution, using the TRACK feature tracking algorithm. Three metrics are defined to characterise the moisture environment of the cyclones: the mean TCWV within a 5 degree radius, which describes the moisture content in the vicinity of precipitation extremes, the minimum within a 9 degree circle, which refers to the conditions in the dry intrusion, and the maximum which refers to the warm conveyor belt. 

The difference between ERA5 and ESA-CCI WV TCWV associated with ETCs is decomposed in the contribution from the better sampling from ERA5, and the bias of ERA5 relative to the ESA-CCI (taken as ground truth). Preliminary results reveal good consistency between ERA5 and ESA-CCI WV estimates for ETCs, highlighting both ERA5 and ESA-CCI as valuable resources for analysing water vapour characteristics in these systems. However, some small regional and seasonally-dependent biases of ERA5 relative to ESA-CCI are identified, and discussed in comparison to other reanalyses and climate model datasets.