Danijel Belušić

A review of recent advances in understanding the meso-and microscale properties of the severe Bora wind

A gusty downslope windstorm that blows at the eastern Adriatic coast is called bora. Similar winds exist at many other places on virtually all continents. Related hourly mean wind speeds surpassing 20 m s −1 , with gusts reaching up to 50 or even 70 m s −1 , in the coastal mountain lee areas are common (hurricane speeds). There has been substantial progress in bora observations and measurements, understanding, modelling and its more detailed prediction during the last 25 yr. It was generally thought before that bora was a falling, mostly thermodynamically driven wind; however, (severe) bora is primarily governed by mountain wave breaking. Understandings of bora interactions and influences on other processes have taken place as well, most notably in the air-sea interaction, but are not completed yet. The overall progress mentioned would not be possible without airborne data, non-linear theory and advances in computational techniques, most notably mesoscale numerical models. Some gaps in bora knowledge are also indicated, for example, dynamical transition from weak to moderate to strong to severe bora flows, where the latter are the main subject here, and vice versa. Moreover, the role of the boundary layer and waves on the upwind side of the bora evolution and the consequent lee side flow structures are inadequately understood; this is especially so for bora at the southern Adriatic coast. The focus here is on stronger bora flows at the NE Adriatic coast.

Atmospheric origin of the devastating coupled air‐sea event in the east Adriatic

Enhanced sea level oscillations with devastating effects occurred on 26/27 June 2003 in the east Adriatic. The phenomenon was recognized as a resonantly coupled air-sea interaction, where the sea wave is forced by the propagating atmospheric pressure disturbance. This study examines the dynamics and predictability of the atmospheric component of the coupled system. First, the initiation, amplification and maintenance of the system are examined. The source of the perturbation is traced back to the Alpine region, where the potentially unstable air in the westerly current lifts over the Alps resulting in convection and wave development. The wave and convective activity apparently couple in a wave-CISK manner and propagate southeastward down the eastern Adriatic coast. Due to the potentially unstable air above the Adriatic, the system additionally amplifies along its track. The dissipation of the system occurs after the landfall. Second, the performance of the numerical model at the resolution that is similar to operational NWP model resolution is examined. This severe event is highly susceptible to the details of the model configuration, presumably due to the sensitivity in coupling of the wave and MCS. Thus although the model is able to reproduce and forecast this kind of event at a fairly coarse resolution, the details of the spatial structure, as well as the time of onset, are questionable. In this case, the greatest sensitivity is primarily in the treatment of moist and convective processes.

Estimation of bora wind gusts using a limited area model

A performance of the wind gust estimate (WGE) method on the bora wind has been examined. Numerical simulations of several bora episodes have been performed using a non-hydrostatic mesoscale model MEMO6. The model captured well the onset and cessation of the bora while the agreement between simulated and observed wind speeds differed from episode to episode. In cases with accurately simulated wind speeds, the WGE results were very good, thus indicating that the method could be used in forecasting bora gusts. The performance of the WGE method for the examined bora cases also suggested a possibility of further simplification of the method for the bora applications. Inspection of the bora flow and its thermodynamical structure revealed that after the bora onset the shear instabilities completely overwhelm the buoyant forces. This means that the parcels with the maximum wind speed in the boundary layer will always be able to reach the surface and result in wind gusts. Therefore, it is enough to have only a vertical profile of the wind speed. Although completely derived from the physical considerations, this represents a very simple way of determining bora wind gusts and can thus be easily implemented in the operational bora wind forecasting models.

Mesoscale dynamics, structure and predictability of a severe Adriatic bora case

The dynamics, structure and temporal evolution of a severe Adriatic bora, which occurred during 14 and 15 November 2004 was inspected. Numerical simulation of the investigated episode was performed by the mesoscale model MM5. The model was validated against the radiosonde data and the wind data from one automatic meteorological station and three ultrasonic anemometers. Two anemometers where located in the region extremely favorable for the bora occurrence (Senj and Vratnik Pass), while the third one was placed in the mainland (Zagreb-Horvatovac). The model reproduced well the onset and the strength of the investigated bora, as well as the establishment of bora-induced potential vorticity (PV) banners. On the other hand, surface turbulent kinetic energy (TKE) was poorly predicted. Inspection of gap wind characteristics indicated the absence of strong dissipation in the flow through a mountain pass, which gives rise to a horizontally elongated jet. Appearance of wave breaking in the lee of a mountain peak leads to the creation of a mountain wake. Shear lines between individual jets and wakes created in this way are then responsible for the generation of PV banners. Also, the ability of the model to predict hourly wind gusts was validated using a recently developed method.

Estimation of length scales from mesoscale networks

This paper reports on the spatial scales of meandering motions using recent data from three mesoscale networks. Although the mesoscale motions extend over a wide range of scales, and time series do not reveal a spectral peak, the examination of the spatial coherence does identify a preferred spatial scale. Several independent methods reveal the same preferred spatial scale for horizontal coherence for a given network. However, the spatial scale differs by a factor of two between the different networks examined here, presumably due to the different topography and surface conditions. The preferred spatial scale increases roughly linearly with range of time scales included in the evaluation. However, the details of this increase do not seem to be predictable, again partly due to site-specific conditions. The preferred spatial scales are of the order of a few kilometres for time scales less than an hour, but may reach tens of kilometres for time scales of several hours. The preferred horizontal length scale determines the area over which the flow features are statistically coherent. Measurements at a single location can be considered as representative for an area comparable to or smaller than the preferred scale.

A note on local and non-local properties of turbulence in the bora flow

On the basis of two-month measurements of the bora wind at Senj, Croatia, with a 1 s temporal resolution, properties of the bora turbulence are inspected using the records of three bora episodes. The spectrum is divided into two parts: high-frequency turbulence (periods less than 1 min) and the low-frequency part (periods between 1 and 10 min) where pulsations appear. We have found that the high-frequency turbulence is generated locally by surface roughness and local wind shear. On the other hand, the low-frequency turbulence, i.e. the pulsations, seems to be independent of the local properties and can therefore be treated as an organized non-local effect. This is in accordance with the studies of the pulsations in the Boulder downslope windstorm.

Grid implementation of the weather research and forecasting model

Atmospheric science is advancing towards very complex phenomena at ever smaller temporal and spatial scales. One of the principal tools utilized in atmospheric science are weather prediction models. These models usually demand large execution times and resource allocation, such as CPU time and storage space. The main goal of our research is porting of the Weather Research and Forecasting model to the Grid infrastructure. Porting has been done through bash scripts that are using existing Grid tools for job and data management, authentification mechanisms, and other application level services produced within the SEEGRID project. In this paper, through a few model runs on the Grid we describe certain benefits not only in the overall execution time but also in the ability of performing concurrent runs of the same model especially for scientific purposes. During the execution, we have also faced some drawbacks in data bandwidth, unreliability of some Grid services and relatively hard control of the model execution flow. The final conclusion is that there is a big need and justification for porting the WRF model to the Grid, although it takes a lot of effort to be properly implemented.

Thermodynamical and Airflow Conditions Within the Lower Troposphere During the Tropopause Fold Event Leading to Elevated Surface Ozone Concentrations

We investigated one episode of a winter stratospheric ozone intrusion over Zagreb, Croatia on February 6, 1990, when unusually high ozone concentrations were measured at two sites in the greater Zagreb area. Lisac et al. (1993) already studied the same episode. They argue that the air must be of stratospheric origin, since photochemistry cannot generate large amounts of ozone at this time of year. Their conclusion is also supported by the spiky character of the ozone level behavior, which was recorded on both stations, and which suggests rapid downward transport related to cross-tropopause exchange (Schuepbach et al., 1999). Lisac et al. confirmed the intrusion by means of analysis of available routine surface and upper air data. Finally, they roughly estimated a probable three-dimensional trajectory of an intruded air parcel. The trajectory starts on February 5 at 01 LST (Zagreb local time) at 300 hPa surface above the Baltic sea, and due to its anticyclonic curvature arrives in Zagreb on February 6 at 13 LST from the east. The above study was limited by the rough spatial and temporal resolution of the routine data, and, it did not offer a clear explanation of the time delay in the ozone peak, which was recorded at one of the measuring sites, which, we believe, is attained in the present study.