Cluster analysis of HYSPLIT backward trajectories for major heatwaves in Spain and Ukraine (1940–2023)

Heatwaves (HWs) are extremely harmful weather phenomena that cause significant damage to the environment and society. Numerous studies have shown a substantial increase in the frequency and intensity of HWs in various parts of the world. Therefore, investigating the meteorological factors contributing to HW formation is an important task. In this study, we investigated the main air transport patterns associated with the most severe HWs observed during 1940-2023 in Spain and Ukraine.

Firstly, based on ERA5 (2 m) air temperature data, we identified all HW events for each grid point in both countries with the heatwaveR package. Following the approach used in many studies, a HW was defined as an extreme weather phenomenon when daily maximum air temperature exceeds 90-th percentile at least for three consecutive days. Additionally, each detected HW episode was categorized as moderate, strong, severe, or extreme based on its maximum observed intensity. A final list of HWs for further analysis in each country was compiled by selecting events with a spatial extent covering more than 20% of the country’s territory and with severe or extreme category identified in at least one grid point. Using this methodology, we selected 80 HW episodes in Spain and 18 in Ukraine.

Backward trajectories for the selected HW episodes were calculated using the HYSPLIT model, with ERA5 3D-data serving as input meteorology. For each HW event, only first three days were considered, regardless of the event's total duration. A starting location for backward trajectories for each HW was defined as a midpoint of its spatial extent. Additionally, to assess the influence of vertical wind shear on trajectory calculations, three altitudes (10, 1500, and 5000 m AGL) were defined as the starting heights. Backward trajectories were initiated hourly over the 3-day period and calculated for the seven days preceding each release hour. In total, 216 backward trajectories were calculated for each HW episode (72 trajectories per release height). The calculated trajectories were then grouped into three clusters based on the HYSPLIT clustering approach and a mean trajectory was determined for each cluster. Along with the cluster analysis, we also identified the dominant circulation types and their evolution during the selected HW episodes. This analysis was performed using the synoptReg package based on ERA5 mean sea level pressure data.

The mean cluster trajectories, calculated for all selected HWs, were used to build trajectory frequency maps, showing the most preferential routes of air masses associated with the severe and extreme HWs. Analysis of these maps revealed that a westerly trajectory flow is the most likely route for air masses responsible for the most intense HWs in Spain and Ukraine. In Spain, air masses are typically transported from the Atlantic, whereas in Ukraine, they traverse across Western Europe. Other directions of air mass transport, including from the south, occur relatively rarely. Our findings align with other similar studies for other regions in Europe, which suggest that heat advection is not dominant mechanism for HW formation.

This work has received funding through the MSCA4Ukraine project, funded by the European Union