Zbyněk Sokol

A diagnostic study of a summer convective precipitation event in the Czech republic using a nonhydrostatic NWP model

Abstract Forecasting, locally, heavy convective precipitation events has been an important topic of investigation for a few decades. It has been recognized that nonhydrostatic models are able to simulate the dynamics of organized convective systems with spatial resolutions of 1 km and less, provided that adequate triggering is entered in the model. Apart from experimental model runs, there are also diagnostic case studies of convective events that employ such models and compare the results with radar data, for example. Nevertheless, the purely deterministic prediction of severe convective systems and the corresponding quantitative precipitation forecast (QPF) is far from being resolved. A research project, focused on case studies of severe convective events in the region of the Czech Republic (CR), has been running since 2000. Data from this study have been used to adapt the nonhydrostatic numerical weather prediction (NWP) Local Model of the German Weather Service (LM DWD) to use a 2.8-km horizontal resolution Small Local Model (SLM) for forecasting and/or warning of severe convection events. The results of several model applications are summarized and discussed. The historical convective event from July 22 to 23, 1998, which caused an extreme precipitation amount and a flash flood over the NE part of the CR, was analyzed with several model runs. The model runs differed in the design of the experiment (starting time of SLM integration, convective parameterization ON/OFF, modification of model orography, etc.). The results of the SLM integration indicate that the inclusion of the convective parameterization smoothes precipitation fields, which is not realistic. The structure of precipitation fields, obtained by the SLM with convective parameterization switched OFF, better corresponds to the spatial structure of radar-derived precipitation. The resulting precipitation depends on the start time of the SLM integration. The results show that the accumulated precipitation does not differ too much, provided the SLM starts its integration at least 6 h before the studied event. The fact that the SLM produces large precipitation in contrast to the routine NWP and that the area precipitation pattern corresponds to the radar precipitation is promising.

Assimilation of radar reflectivity into the LM COSMO model with a high horizontal resolution

An assimilation of radar reflectivity into a numerical weather prediction (NWP) model with a horizontal resolution of 2.8 km is presented and applied to three severe convective events. The suggested assimilation method takes into account differences between the model and radar-derived precipitation in modifying vertical profiles of water vapour mixing ratio in each model time step by the nudging approach. Version 3.9 of the LM COSMO (Local Model COSMO) –NWP model used in this study includes the explicit formulation of the cloud and rain processes involved. Two variants of the assimilation technique are designed and outputs of their implementation are compared. The first variant makes use of the ground data only, while the second utilises vertical profiles of precipitation water. Both variants provide an improvement of precipitation forecast in comparison with outputs of the control run without assimilation procedures applied. When the assimilated radar data indicate initial precipitation near an expected storm, the NWP model is capable of forecasting basic features of the storm development two to three hours ahead. Three case studies are presented. In one, the assimilation method that takes into account the vertical structure of the precipitation water yields better results than the others which utilise ground data only. However, for the remaining two case studies both types of the assimilation method produce comparable results. Copyright

Improvement of Local Categorical Precipitation Forecasts from an NWP Model by Various Statistical Postprocessing Methods

Several statistical postprocessing methods are applied to results from a numerical weather prediction (NWP) model to test the potential for increasing the accuracy of its local precipitation forecasts. Categorical (Yes/No) forecasts for 12hr precipitation sums equalling or exceeding 0.1, 2.0 and 5.0 mm are selected for improvement. The two 12hr periods 0600-1800 UTC and 1800-0600 UTC are treated separately based on NWP model initial times 0000 UTC and 1200 UTC, respectively. Input data are taken from three successive summer seasons, April-September, 1994-96. The forecasts are prepared and verified for five synoptic stations, four located in the western Czech Republic, and one in Germany near the Czech-German border.Two approaches to statistical postprocessing are tested. The first uses Model Output Statistics (MOS) and the second modifies the MOS approach by applying a successive learning technique (SLT). For each approach several statistical models for the relationship between NWP model predictors and predictand were studied. An independent data set is used for forecast verification with the skill measured by a True Skill Score.The results of the statistical postprocessing are compared with the direct model precipitation forecasts from gridpoints nearest the stations, and they show that both postprocessing approaches provide substantially better forecasts than the direct NWP model output. The relative improvement increases with increasing precipitation amount and there is no significant difference in performance between the two 12hr periods. The skill of the SLT does not depend significantly on the size of the initial learning sample, but its results are nevertheless comparable with the results obtained from the MOS approach, which requires larger developmental samples.

Skill of 6HR and 12HR thunderstorm forecasts based on assessment of larger-scale information

SummaryThe relationship between information, contained in aerological data from the European area, and a thunderstorm occurrence in the area of the Czech Republic was investigated with input data from the period of May–September 1989–1991. SYNOP reports from Czech ground stations were utilized to assess event occurrence. TEMP 00UTC and TEMP 12UTC reports from European stations were used to determine potential diagnostic predictors, and the TEMP00 data served as the input data set for the 12hr mesoscale model forecast to gain prognostic predictors. Each of the two diagnostic data sets from 00UTC and 12UTC and of the prognostic data set comprised about 400 predictors/predictand elements. The categorical forecast of thunderstorm occurrence, based on the application of linear regression and a simple version of pattern recognition, is discussed. The critical success index was determined for every type of forecast and used to assess forecast skill.

The inclusion of surface synoptic data in very short range predictions of convective events

SummaryPredictor vectors, including upper air as well as surface data, were used for categorical forecasting convective events over a subregion of the Czech territory, and the effect of including surface variables in the predictor vector was examined. While upper air data were considered as Perfect Prognosis, the surface data were successively included according to the time of their origin. The forecasting technique was based on linear multiple regression with learning, and the accuracy of the forecast was measured by the Critical Success Index. The input data from the three May-September periods in 1989–91 were used, and the first year served as the learning set. The aerological data from TEMP 12 UTC, simulating Perfect Prognosis, were the source of the upper air predictors. The performance of all, upper air, surface and combined, predictors were evaluated and compared. It turned out that the improvement of prediction accuracy due to the inclusion of surface variables was not negligible. Significant improvements were made in the forecasts of thunderstorm occurrence between 18 and 24 UTC.

Simulations of the Impact of Lake Area on Local Microclimate by Using COSMO NWP Model

A hydric coal mine restoration creates new water areas, which change surface characteristics of the locality. These changes consist primarily in different thermal properties, different surface roughness and different albedo compared to the original surface. It is evident that the hydric restoration influences the atmosphere and thus affects the temperature and humidity, as well as other meteorological quantities around the lake. The impact of the restoration can influence much larger areas by changing precipitation climatology.

Tests of data assimilation and data initialization for the purposes of short-range weather forecast

SummaryThe initialization and assimilation of input data were studied and tested by the adiabatic version of a simple numerical model for short-range weather forecast.The initialization was based on the utilization of a digital filter technique. The method succeeded in removing high-frequency oscillations from prognostic pressure fields. However, excessive smoothing deteriorated the accuracy of the prediction at the lowest levels of the atmosphere.The data assimilation was performed using the nudging method. Three versions of the nudging method in a splitting scheme were tested. The inclusion of the assimilation at the end of the integration step proved to be the best. The assimilation damped the oscillations of prognostic surface pressure fields and slightly improved the pressure prediction at the lowest levels of the atmosphere.