Forecast errors attributed to synoptic features and the role of diabatic heating for extratropical cyclones

It is often argued that numerical weather prediction models remain deficient in forecasting specific weather features and that such deficiencies contribute significantly to overall forecast errors. To clarify these claims, we quantify how extratropical cyclones (ETCs), fronts, upper tropospheric jets, moisture transport axes (MTAs), and cold-air outbreaks (CAOs) contribute to short-term (12-h) forecast errors and biases in the ERA5 reanalysis dataset from 1979 to 2022. Employing a feature-based attribution method, we evaluate errors globally, focusing particularly on temperature, moisture, and wind fields, and examine regional and seasonal variations during winter (DJF) and summer (JJA). The presence of weather features is generally associated with increased forecast errors (RMSEs) compared to feature-free conditions. RMSEs are especially pronounced for moisture fields in conjunction with fronts and MTAs, where errors in total column water vapor can be twice as large. ETC-related errors are more pronounced in the low-level wind field. During CAOs, on the other hand, errors are reduced. In terms of systematic biases, wind speeds and moisture are underestimated along western boundary currents, together with insufficient moisture transport along MTAs.

Given that ETCs are the most notable example, where forecasts provide less added value in most cases we also employ a cyclone-centred composite framework for North Atlantic wintertime (DJF) ETCs using the ERA5 reanalysis for the period 1979 to 2022. ETCs are categorised into strong and weak diabatic heating at the time of their maximum intensification. While both groups exhibit a systematic underestimation of cyclone intensity, the error structures are markedly distinct. The weak heating group is characterised by an intensity underestimation near the cyclone core, whereas the strong heating group features a pronounced southwestward displacement bias together with a domain-wide intensity underestimation. After removing the displacement bias, the strong heating group reveals an overestimation of low-level winds within the cold conveyor belt, sting jet, and dry intrusion regions, but a clear underestimation of moisture transport in the warm sector. These biases are accompanied by a pronounced overestimation of 850 hPa kinematic frontogenesis near the centre, likely associated with the wind field errors, and a substantial overestimation of total column liquid water along the bent-back warm front. This overestimated liquid water is likely related to the stronger frontogenesis, which induces an over-intensified secondary circulation. In contrast, cyclones in the weak heating group exhibit an underestimation of wind speed and moisture near the centre, consistent with the near centre intensity underestimation. Our findings highlight the impact of diabatic heating on structural cyclone forecast biases that can guide future model improvements.