Are Trends in Convective Parameters over the United States and Europe Consistent between Reanalyses and Observations?

In this work, long-term trends in convective parameters are compared between ERA5, MERRA-2, and observed rawinsonde profiles over Europe and the United States including surrounding areas. A 39-yr record (1980-2018) with 2.07 million quality-controlled measurements from 84 stations at 0000 and 1200 UTC is used for the comparison, along with collocated reanalysis profiles. Overall, reanalyses provide signals that are similar to observations, but ERA5 features lower biases. Over Europe, agreement in the trend signal between rawinsondes and the reanalyses is better, particularly with respect to instability (lifted index), low-level moisture (mixing ratio), and 0-3-km lapse rates as compared with mixed trends in the United States. However, consistent signals for all three datasets and both domains are found for robust increases in convective inhibition (CIN), downdraft CAPE (DCAPE), and decreases in mean 0-4-km relative humidity. Despite differing trends between continents, the reanalyses capture well changes in 0-6-km wind shear and 1-3-km mean wind with modest increases in the United States and decreases in Europe. However, these changes are mostly insignificant. All datasets indicate consistent warming of almost the entire tropospheric profile, which over Europe is the fastest near ground whereas across the Great Plains it is generally between 2 and 3 km above ground level, thus contributing to increases in CIN. Results of this work show the importance of intercomparing trends between various datasets, as the limitations associated with one reanalysis or observations may lead to uncertainties and lower our confidence in how parameters are changing over time.

Automated summary

This study compares long-term trends in convective parameters between ERA5, MERRA-2, and observed rawinsonde profiles over Europe and the United States. The results show that the reanalyses provide signals that are similar to the observations, but ERA5 has lower biases. The trend agreement between rawinsondes and reanalyses is better in Europe, particularly for instability, low-level moisture, and 0-3-km lapse rates. However, robust increases in convective inhibition, downdraft CAPE, and decreases in mean 0-4-km relative humidity are consistent signals for all three datasets and both domains. The study also highlights the importance of comparing trends between different datasets to reduce uncertainties and increase confidence in parameter changes over time.