Vol. 121, No. 2 * Pages 101–208 * April - June 2017

In April 2015, a serious industrial pollution gained large public interest in Budapest, the capital city of Hungary. An abandoned industrial site of the former Budapest Chemical Works company was found to contain 2000–3000 tonnes of leaking industrial waste and dangerous chemicals. As the former factory is located in a densely populated urban environment, serious public health concerns have risen. The pollution has been transported for several years from the industrial site to the neighboring areas by two ways: with groundwater transport and in the atmosphere.
This study attempts to estimate the main characteristics of atmospheric transport of pollutants originating from the Budapest Chemical Works site. The 13-year long dispersion pattern is investigated using a Gaussian dispersion model, taking into account the strong weather dependence of emission rates through deflation and evaporation. The very limited information of the amount, composition, and temporal evolution of the leaking chemicals makes it impossible to provide emission estimates; however, the spatial distribution and temporal characteristics of the pollution can be investigated.
Weather-dependent emission rate was found to be the dominant factor of the pollution, counteracted by the atmospheric stability. Largest concentrations were present in the spring and summer and during the day, while nighttime emissions were generally weak. The main direction of the dispersion was towards the south-east, however, deflating and evaporating materials showed largely different results.

The paper discusses the long-term variability of maximum (Tmax) and minimum (Tmin) air temperature variations, both occurring from one day to the next and over several consecutive days (3-4), in Krakow (Poland, Central Europe) from 1826 to 2015 (i.e., over a period of 190 years). The authors analyzed the seasonal variability of short-term variations in air temperature, looking at the most significant changes (±10 °C), as well as at their dynamics and trends over the analyzed multi-annual period. A clear decrease has been observed both in the values of short-term Tmax and Tmin variations and in the number of cases with their significant fluctuations. The decrease has been gradual, without clear abrupt changes to the overall trend. The greatest short-term variations in temperature were most frequent in the cooler half-year, being smaller and less frequent in the summer months. If the observed trend persists, in the upcoming years we can expect a further decrease in the dynamics of variations in thermal conditions, i.e., short-term variations in temperature may more frequently be small, i.e., ±0.1–4.0 °C. However, it is worth noting that Tmax more frequently increased from one day to the next and over several consecutive days, while Tmin more frequently decreased. The reasons for the analyzed changes remain unclear. It seems that natural factors, mainly including the advection of air masses, have a significant impact on short-term variations in air temperature, coupled with local factors, which have been strengthened by the human impact on the environment, including the urban heat island.

In order to efficiently integrate renewable energy sources – the production of which can be planned harder – in the energy grid, quite accurate forecasts are needed. Especially in Hungary, where energy storage is yet an unsolved problem. The limited ability of wind forecast means that the delivered power of the wind farms cannot be predicted with sufficient accuracy. This work focuses on the improvement of wind power forecast precision by using different statistical methods. In the first part of the paper, simple BIAS correction approaches and more complex ensemble based methods are applied to improve the power prediction for the whole country. The second part of the work focuses on enhancement of the wind power forecast of a single wind farm. While mostly only wind speed is taken into consideration for the calculation of the generated power, it is shown that air density is also an important factor in the equations. Autoregressive filtering is launched in order to show that wind speed and power forecasts can also be improved by this kind of statistical method.

This study presents a methodology to assess the climate change impacts on wind conditions and wind energy potential on multiple levels near the surface over the Carpathian Basin and Hungary. The methodology is based on ALADIN-Climate regional climate model results and ERA-Interim re-analysis data.
Since wind energy estimations require wind data in specific hub (turbine) heights, in addition to the 10-meter standard, we evaluate wind speed on 50, 75, 100, 125, and 150 meters above the surface to cover the range of most frequently used hub heights. The main concept of the method is to compute the wind velocity on these levels directly from data on the neighboring model levels instead of extrapolating from the 10-meter wind speed applying a wind profile. Besides giving more accurate velocity values, the use of multiple levels allows us to examine the changes in the vertical profile of near-surface winds as well.
The model results are validated with ERA-Interim re-analysis for the 1981–2000 period. Despite a systematic negative bias, ALADIN-Climate reproduces the main wind characteristics in the Carpathian Basin reasonably. The future projection was carried out considering the RCP8.5 emission scenario and was evaluated for the 2021–2050 and
2071–2100 periods. The projection results show a mild future increase in the average wind speed over most parts of the integration domain. The changes over Hungary are more prominent in 2021–2050 with a slight but statistically significant 7% annual increase. The mean annual change in potential power has similar characteristics, only with higher, 8–13% growth.
As our aim is the demonstration of a methodology, our investigation is based on the outputs of a single climate model simulation, however, to provide some hints about projection uncertainties, we compared our future estimates with further studies which confirmed our main conclusions.

Global radiation output fluxes predicted by numerical weather forecast model AROME were verified by using measured high accuracy global radiation data from the 19 most reliable network stations of the Hungarian Meteorological Service. Three suitably-selected months (April, June, August) from 2013 were used for the study. Differences between observed and forecasted values were analyzed separately for all cases, overcast cases, and cloudless (clear-sky) cases. It was found that AROME performs well for clear cases, and its goodness decreases as cloudiness increases. For cloudless cases, using aerosol optical depth, graybody optical depth, and relative global radiation to represent radiative transmission condition of the atmosphere, it was found that AROME overestimates atmospheric radiation transmission for cases of high turbidity and underestimates it for very clear conditions. It means that radiation transmission scale of the atmosphere produced by the model is more narrow than that of true atmosphere.