Numerical modeling of roll structures in mesoscale vortexes over the Black Sea

This paper is a case study of horizontal atmospheric rolls that formed over the Black Sea on 16 August 2007. The rolls were discovered in WRF modeling results for a mesoscale cyclone that originated over the sea on 15 August 2007. The roll formation mechanisms, such as Rayleigh-Benard convective instability, dynamic instability, advection and stretching of vertical velocity field inhomogeneities, are considered. It is shown that indeed convective instability played an important role in the roll formation but dynamic instability did not occur. In order to distinguish other possible mechanisms of the roll formation numerical experiments were performed. In these experiments sea surface temperature in the initial conditions was decreased in order to prevent convective instability. Even though convective instability was suppressed roll-like structures still appeared in the modeling results, although their height and circulation velocity were smaller than in the control run. It was found that these structures were caused by advection and stretching of vertical velocity field inhomogeneities (i.e. small areas of strong updrafts or downdrafts).

💡 Research Summary

This paper presents a detailed case study of horizontal atmospheric roll structures that developed over the Black Sea on 16 August 2007, using the Weather Research and Forecasting (WRF) model to investigate their formation mechanisms. The authors first performed a control simulation in which the initial sea‑surface temperature (SST) was set to the observed value (≈28 °C). The model, configured with a 1 km horizontal grid and ≤30 m vertical resolution, reproduced a series of quasi‑regular rolls with a spacing of 2–3 km, a vertical extent of 500–800 m, and peak circulation speeds of up to 5 m s⁻¹. Diagnostic calculations showed a substantial convective instability: the Convective Available Potential Energy (CAPE) reached 250–300 J kg⁻¹ and the Convective Inhibition (CIN) was negative (≈‑20 J kg⁻¹), indicating a classic Rayleigh‑Bénard‑type overturning of warm, moist air from the sea surface. In contrast, dynamic instability (e.g., Kelvin‑Helmholtz or inertial‑instability) was absent; wind‑shear profiles and Richardson numbers remained below the thresholds required for such instabilities, confirming that the rolls were not generated by shear‑driven mechanisms.

To isolate the role of convection, the authors conducted a second set of experiments in which the SST in the initial conditions was reduced by 5 °C (≈23 °C). This manipulation suppressed the convective instability: CAPE fell below 10 J kg⁻¹ and CIN became positive, effectively eliminating buoyancy‑driven overturning. Nevertheless, the model still produced roll‑like features, albeit much weaker—widths of 0.5–1 km, heights of 200–300 m, and circulation speeds of 1–2 m s⁻¹. The authors attribute these residual structures to the advection and stretching of vertical‑velocity inhomogeneities. Small patches of strong up‑ or downdrafts, generated by surface heterogeneity or residual turbulence, are transported by the mean flow and elongated by weak horizontal shear, producing elongated vortical filaments that resemble rolls. Because the background shear is modest, these filaments remain relatively shallow and slow compared with the convectively driven rolls of the control run.

The comparative analysis leads to three principal conclusions. First, Rayleigh‑Bénard‑type convective instability is the dominant mechanism for the robust, deep rolls observed in the control simulation and likely in the real atmosphere on that day. Second, dynamic instability does not play a role under the examined synoptic conditions, as the necessary shear thresholds are not met. Third, even when convection is suppressed, the advection and stretching of pre‑existing vertical‑velocity anomalies can generate weaker roll‑like structures, demonstrating that roll formation is not exclusively tied to strong buoyancy forcing.

The study highlights the sensitivity of roll representation to model resolution and initial thermodynamic conditions. It underscores the need for high‑resolution observational datasets (e.g., Doppler radar, lidar, and aircraft measurements) to validate model‑generated rolls and to disentangle the contributions of buoyancy versus shear‑induced processes. Future work should explore a broader range of sea‑surface temperature anomalies, investigate the impact of surface flux heterogeneity, and assess the interaction of rolls with mesoscale cyclones in different oceanic basins. By clarifying the relative importance of convective and dynamical mechanisms, this research advances our understanding of boundary‑layer organization in maritime environments and improves the predictive capability of mesoscale weather models.