Conventional air quality monitoring networks typically tend to be sparse over areas of interest. Because of the high cost of establishing such monitoring systems, some areas are often completely left out of regulatory monitoring networks. Recently, a new paradigm in monitoring has emerged that utilizes low-cost air pollution sensors, thus making it possible to reduce the knowledge gap in air pollution levels for areas not covered by regulatory monitoring networks and increase the spatial resolution of monitoring in others. The benefits of such networks for the community are almost self-evident since information about the level of air pollution can be transmitted in real time and the data can be analysed immediately over the wider area. However, the accuracy and reliability of newly produced data must also be taken into account in order to be able to correctly interpret the results. In this study, we analyse particulate matter pollution data from a large network of low-cost particulate matter monitors that was deployed and placed in outdoor spaces in schools in central and western Serbia under the Schools for Better Air Quality UNICEF pilot initiative in the period from April 2022 to June 2023. The network consisted of 129 devices in 15 municipalities, with 11 of the municipalities having such extensive real-time measurements of particulate matter concentration for the first time. The analysis showed that the maximum concentrations of PM2.5 and PM10 were in the winter months (heating season), while during the summer months (non-heating season), the concentrations were several times lower. Also, in some municipalities, the maximum values and number of daily exceedances of PM10 (50 μg/m3) were much higher than in the others because of diversity and differences in the low-cost sensor sampling sites. The particulate matter mass daily concentrations obtained by low-cost sensors were analysed and also classified according to the European AQI (air quality index) applied to low-cost sensor data. This study confirmed that the large network of low-cost air pollution sensors can be useful in providing real-time information and warnings about higher pollution days and episodes, particularly in situations where there is a lack of local or national regulatory monitoring stations in the area.
1 January 2020 in Nuclear Technology and Radiation Protection
In nuclear physics experiments, it is very important to isolate the measured quantities from electromagnetic noise. Without this possibility, it is impossible to obtain usable experimental results since natural electromagnetic noise can be several orders of magnitude larger than the measured magnitude. In order to enable such measurements, it is necessary to eliminate electromagnetic noise from the experimental procedure. This is achieved by shielding against electromagnetic radiation. In this paper, experiments were performed to protect a room from electromagnetic noise. By applying all known methods of shielding against electromagnetic radiation, it was concluded that the room can be protected from the electrical component, but it is impossible to protect it from the magnetic component of electromagnetic radiation.
Mathematical methods are the basis of most models that describe the natural phenomena around us. However, the well-known conventional mathematical models for atmospheric modeling have some limitations. Machine learning with Big Data is also based on mathematics but offers a new approach for modeling. There are two methodologies to develop deep learning models for spatio-temporal image prediction. On these bases, two models were built—ConvLSTM and CNN-LSTM—with two types of predictions, i.e., sequence-to-sequence and sequence-to-one, in order to forecast Aerosol Optical Thickness sequences. The input dataset for training was NASA satellite imagery MODAL2_E_AER_OD from Terra/MODIS satellites, which presents global Aerosol Optical Thickness with an 8 day temporal resolution from 2000 to the present. The obtained results show that the ConvLSTM sequence-to-one model had the lowest RMSE error and the highest Cosine Similarity value. The advantages of the developed DL models are that they can be executed in milliseconds on a PC, can be used for global-scale Earth observations, and can serve as tracers to study how the Earth’s atmosphere moves. The developed models can be used as transfer learning for similar image time-series forecasting models.
The attention mechanism in natural language processing and self-attention mechanism in vision transformers improved many deep learning models. An implementation of the self-attention mechanism with the previously developed ConvLSTM sequence-to-one model was done in order to make a comparative evaluation with statistical testing. First, the new ConvLSTM sequence-to-one model with a self-attention mechanism was developed and then the self-attention layer was removed in order to make comparison. The hyperparameters optimization process was conducted by grid search for integer and string type parameters, and with particle swarm optimization for float type parameters. A cross validation technique was used for better evaluating models with a predefined ratio of train-validation-test subsets. Both models with and without a self-attention layer passed defined evaluation criteria that means that models are able to generate the image of the global aerosol thickness and able to find patterns for changes in the time domain. The model obtained by an ablation study on the self-attention layer achieved better outcomes for Root Mean Square Error and Euclidean Distance in regards to developed ConvLSTM-SA model. As part of the statistical test, a Kruskal–Wallis H Test was done since it was determined that the data did not belong to the normal distribution and the obtained results showed that both models, with and without the SA layer, predict similar images with patterns at the pixel level to the original dataset. However, the model without the SA layer was more similar to the original dataset especially in the time domain at the pixel level. Based on the comparative evaluation with statistical testing, it was concluded that the developed ConvLSTM-SA model better predicts without an SA layer.
1 January 2021 in Nuclear Technology and Radiation Protection
The paper considers the application of nuclear magnetic resonance for measurement of fluid-flow. The paper is of an experimental nature. Flowmeter based on nuclear magnetic resonance is extremely precise. The combined measurement uncertainty can be 0.1 %. Such a value of measurement uncertainty indicates that it is a matter of a deterministic and not of a stochastic quantity. This high degree of reliability of the method is theoretically and mathematically described. The paper presents a measurement scheme for flow measurement. Water flow measurement was performed on the principle of nuclear magnetic resonance and on the basis of tritiated water (which is considered to be the most accurate classical method). The obtained results show that the measurement of flow based on nuclear magnetic resonance is more accurate (especially at higher flow). This is explained by the higher inertial mass of HTO tritiated water molecules than the standard H2 O mass and the possible transition of tritiated water to H3HeO. In this way, it has been proven that tracing water based on nuclear magnetic resonance is the only real tracing of water by water. The obtained results show that tracing water with tritiated or heavy water is not tracing water by water which is explained by different inertial masses.
1 January 2022 in Nuclear Technology and Radiation Protection
The research presents an experiment with a model of an electronic generator for energy injection into the plasma of a fusion reactor. By recording a non-ionizing field in the vicinity of a fusion reactor, it was determined that this field has an extremely high growth rate. At the site of the maximum intensity of the field of non-ionizing radiation, commercial surge arresters with a flexible model of surge arresters were used for experimentation. It has been found that the commercial surge arresters have an efficiency of about 20%. For the efficiency of the flexible model, it was found to be slightly less than 40% (and to be achieved by the application of alpha particle radiation). Since neither of these efficiencies guarantee reliable operation of the gas-filled surge arrester, it was concluded that essential electronics in the vicinity of the fusion generator must be protected. However, since this protection can only be implemented in a fusion reactor, the fact remains that the environment of such a reactor is extremely contaminated with non-ionizing radiation. Commercial surge voltages are isolated for testing since the protection of electronic circuits from fast overvoltages is a critical point for the functioning of modern electronics.