B. Ya. Shmerlin’s research while affiliated with Scientific and Production Association "Typhoon" and other places

The automated forecasting of waterspouts off the Black Sea coast is considered. The technique for the short-range forecasting of waterspout-risk situations and for the detection of waterspout-risk areas based on weather radar data in real time are presented. The results of the author’s testing of the technique for the period from August 2017 to October 2018 are given. A possibility of the automated preparation of waterspout warnings is substantiated. The prospects of using the experience gained for the development of the automated tornado forecasting system for the European part of Russia are analyzed.

In many studies indices of convective instability (hereinafter simply indices) are used to analyze and predict tornado-dangerous situations. For calculating the meteorological fields from which indices were subsequently calculated, the WRF-ARW version 3.4 was used – the non-hydrostatic, regional weather forecasting system. In the works (Novitskii et al, 2016; 2018) as an example of the calculation of 10 tornadoes that occurred at different times in the European territory of the Russian Federation, we show that the most informative from the point of view of forecasting tornado-dangerous situations and providing a minimum number of false warnings is the STP (significant tornado parameter) index. The characteristic time, during which STP index exceeds threshold value, is within the order of an hour, the size of the regions of localization of the values of the indices above the threshold is within the order of several tens kilometers. We proposed along with the STP index to involve the vertical velocity field, calculated in the WRF model, in the analysis and forecast of tornado-dangerous situations. We show that the value of the STP index above the threshold leads within the WRF model to the formation of a localized intense convective cell in the vertical velocity field in the vicinity of the maximum value of the index and at the moment of reaching this value. The possibility of using the STP index to predict tornado-dangerous situations with a lead time of up to three days with an accuracy of 150 km in space and several hours in time is demonstrated. A new approach to short-term forecasting of tornadoes is proposed. It is based on calculating the fields that are visible on the radar screen, using the WRF model forecast. Such fields are the fields of maximum radar reflectivity, upper cloud boundary and integral vertical water content. The comparison of the prognostic fields with the real fields that the radar sees allows us to specify a real convective system at the time of its formation in which the STP index will subsequently reach a threshold value and a tornado will appear. This can enlarge a lead time of tornado warnings to several hours, which currently averages 13 minutes. The approach can also be used for forecasting other dangerous convective phenomena, as well as in any other forecast models for current forecast correction by using incoming radar (satellite) information.

Calculation of convective indices using the WRF model for the analysis and forecasting of tornado-risk situations involved the calculation of vertical velocity field. The results of calculations for four tornados registered in 2015 are presented. It is shown that if the values of the indices are above the thresholds, the localized intensive convective cell is formed in the model in the vicinity of the maximum values of the indices and at the moment when these values are reached. The possibility of using this result as an additional predictor of tornado occurrence is discussed. It confirms the conclusion on the possibility of prediction of tornado-risk situations with the lead time to three days and with the accuracy to several hours in time and to 200 km in space.

The problem of convective instability of an atmospheric layer containing a horizontally finite region filled with a cloud medium is considered. Solutions exponentially growing with time, i.e., solitary cloud rolls or spatially localized systems of cloud rolls, have been constructed. In the case of axial symmetry, their analogs are convective vortices with both ascending and descending motions on the axis and cloud clusters with ring-shaped convective structures. Depending on the anisotropy of turbulent exchange, the scale of vortices changes from the tornado scale to the scale of tropical cyclones. The solutions with descending motions on the axis can correspond to the formation of a tornado funnel or a hurricane eye in tropical cyclones.

The possibility of advance tornado warning is analyzed. The tornado observed in Bashkortostan on August 29, 2014 is considered as an example. To compute meteorological fields, the WRF model with high spatiotemporal resolution is used. Indices of convective instability are calculated. The analysis of variations in indices enabled forecasting the tornado generation with the lead time up to three days and with the accuracy up to several hours in time and 200 km in space. The possibility is demonstrated of registering and nowcasting tornados by using the available software for radar data processing. The polential is discussed of the joint use of such information for developing the syslem of monitoring and forecasting of severe weather events including tornados.

The classical Rayleigh theory of convective instability of a viscous and heat conductive rotating atmospheric layer is generalized to the case of phase transitions of water vapor both for the precipitation convection (PC) and for the nonprecipitation (NPC) one. A principal difference is stated between moist convection and Rayleigh convection, on the one hand, and PC and NPC, on the other hand. In particular, the instability region on the plane of model parameters turned out to generally consist of two subregions, in one of which the localized axisymmetric disturbances with a tropical cyclone (hurricane) structure have the highest growth rate. In case of PC the ascending motions on the axis of symmetry correspond to such disturbances, in case of NPC a spontaneous growth of localized vortices both with ascending and descending motions on the axis is possible. Under other parameters values in case of PC spatially periodic cloud structures (convective rolls or closed cloud cells) have the highest growth rate and in case of NPC-mesoscale systems of convective rolls or mesoscale cloud clusters with annular cloud structures.

Within the framework of the hydromechanical model (HMM), proposed by one of the authors, a tropical cyclone (TC) motion is defined by a largescale wind field and a TC intensity. The model contains parameters describing TC and its interaction with wind field. The diagnostic, quasi-prognostic and prognostic calculations of TC movement are carried out. Diagnostic and quasi-prognostic calculations mean that an objective analysis of a large scale wind field and an objective analysis of a TC intensity is used during a TC whole lifetime. In case of diagnostic calculations, model parameters (constants for each TC) are defined from the best coincidence between the real and calculated track of a TC during a TC whole lifetime; for quasiprognostic calculations they are defined during the preliminary "preprognostic" period. Diagnostic calculations show that the HMMrather correctly describes peculiarities of a TC motion. Quasi-prognostic calculations show that model parameters may be rather correctly defined during a preliminary "preprognostic" period. The results of the diagnostic, quasi-prognostic and prognostic calculations are presented.

Analyzed are atmospheric conditions at the tornado formation in Obninsk (the Kaluga oblast) on May 23, 2013. The sets of meteorological parameters obtained from different sources including the high-altitude meteorological mast in Obninsk and giving the full pattern of conditions in the lower 300-meter atmospheric layer are used as initial data. Presented are the results of computation of some indices of convective instability using the WRF model with the high spatial resolution as applied to the conditions in Obninsk on May 23, 2013. Given is the assessment of the possibility of the advance forecasting of tornado-risk conditions based on the mentioned approach.

The classical Rayleigh problem of convective instability is generalized to the case of water vapor condensation in the atmosphere. We present an analytical solution demonstrating a fundamental difference between moist convection and Rayleigh convection: the curve of the critical Rayleigh number versus the number characterizing the intensity of condensation heat release consists of two parts, with spatially localized neutral solutions corresponding to one of them. Spatially periodic neutral solutions correspond to the second part of the curve; these are characterized by a significant localization of the regions of ascending motions. The theory describes the nucleation and development of individual convective clouds and ordered cloud structures.

Convective motions in moist saturated air are accompanied by the release of latent heat of condensation. Taking this effect into account, we consider the problem of convective instability of a moist saturated air layer, generalizing the formulation of the classical Rayleigh problem. An analytic solution demonstrating the fundamental difference between moist convection and Rayleigh convection is obtained. Upon losing stability in the two-dimensional case, localized convective rolls or spatially periodic chains of rollers with localized areas of upward motion evolve. In the case of axial symmetry, the growth of localized convective vortices with circulation characteristic of tropical cyclones (hurricanes) is possible at the early stages of development and on the scale of tornados to tropical cyclones.