Floristic composition in three industrial areas with soils contaminated by heavy metals (As, Cd, Cu, Hg, Pb, Zn) and organic pollutants (polychlorinated biphenyls) was studied. The content of Pb was only significantly correlated with the floristic composition and explained 13.8% of its variability considering spatial dependency of the sites. No correlation was found for PCBs. Altogether, 237 plant vascular species were found at three study sites (117, 133 and 105, respectively). The three study areas differed in their species composition represented by their own characteristic species. The gradient in the content of natives/non-natives, species number, prevailing life forms and indicator values for plant species investigated was revealed. Based on our results, for phytoremediation purposes we can select productive plant species with high biomass and ability to accumulate large amounts of heavy metals or organic compounds and surviving on soils with low mineral content.

No abstract

<p>It is generally accepted that weather forecasts contain errors due to the chaotic nature of the atmosphere. Regression models, such as neural networks, are traditionally trained to minimize the pixel-wise difference between their predictions and ground truth. The major shortcoming of these models is that they express uncertainty about prediction with blurring, especially for longer prediction lead times. One way to tackle this issue is to use a generative adversarial network, which learns what real precipitation should look like during training. Coupled with a loss, such as Mean Squared or Mean Absolute Error, these networks can produce highly accurate and realistic nowcasts. As there is an inherent randomness in those networks, they allow to be sampled from, just like ensemble models, and various probabilistic metrics can be calculated from the samples. In this work, we have designed a physically-constrained generative adversarial network for radar reflectivity prediction. We compare this network to one without physical restraints and show that it predicts events with higher accuracy and shows much less variance among its samples. Furthermore, we explore fine-tuning the network to the prediction of severe weather events, as an accurate prediction of these benefits both automated warning systems and forecasters.</p>

<p>A Start-Up company (Meteosense, a subsidiary of Meteopress, Czech republic, and Idokep, Hungary)  in collaboration with a National Meteorological Office in Austria (ZAMG) is preparing and deploying a radar network consisting of affordable X-band weather radars. With a minimum team of three people and based on Lean methodology (Build-Measure-Learn) the plan was set-up with milestones along the way. </p><p>The presentation will describe Stage Zero - radar site selection criteria, planning and simulating, Stage One - upgrading existing radar in Vienna to 2.4 meters antenna and Stage Two - planning and deployment of second radar in Austria. Currently, Stage Two is on the way. Stages Three and further will be also briefly described and mistakes and lessons learned will be revealed. </p><p>In Stage Zero we will describe how we chose locations for new radars and how we plan the expansion of our radar network in Austria with the help of our radar simulators and experience from building weather radar network in the Czech Republic, Slovakia, Hungary and Croatia.</p><p>In Stage One we will describe how we upgraded previously installed radar in Vienna from 1.2-meters antenna to 2.4-meters antenna. We will show the problems that occurred during installation and lessons learned will be revealed. We will also show the results and data from this radar.</p><p>In Stage Two we will describe the planning and deployment of our second radar in Austria and this stage will also include how we plan and prepare radar installation in general.</p><p>Stages Three and further will include our future plans in Austria.</p>

No abstract