81 research outputs found
PM2.5 dispersion in Venice area: a model validation
No full textA multidisciplinary project was developed with the aim of better understand PM2.5 primary sources and secondary aerosol formation and compositions. A model system was used to simulate four periods during different seasons in 2009 for which both organic and inorganic measured data were available. Input data were estimated and formatted as requested by models. Measured and predicted data were compared for the three stations and during different seasons in order to test model performance
A Big Data Architecture for Learning-Based Source-Term Estimation
No full textAfter the detection of a radioactive substance of unknown origin in the atmosphere, the source location is estimated via inverse modelling. Depending on various factors, such as the spatial resolution desired, traditional inverse modelling can be computationally time-consuming and therefore its application can be problematic when timing is critical. In a complementary presentation (S. Andronopoulos et al., “Towards Inverse Source Term Estimation using Big Data Technologies”), we discuss a data- scientific approach to source term estimation, which allows us to perform the bulk of the processing prior to such an event taking place, therefore allowing for rapid estimation. This poster presents the big data platform that enabled the implementation of this work and the design and rationale behind our software architecture
Overview of WP4: extension of atmospheric dispersion and consequence modelling in Decision Support Systems
No full textThe activities that have been carried out within the frame of Working Package 4 of PREPARE project, entitled “Extension of atmospheric dispersion and consequence modelling in Decision Support Systems (DSSs)” can be grouped under the following themes: (1) Source term (quantity and time variation of release rate of radionuclides from a nuclear accident) estimation based on optimal combination of atmospheric dispersion modelling with measurements through simple and more advanced computational methods integrated within operational DSSs. (2) Implementation of physicochemical properties of radionuclides emitted as particulate matter in computational modules of DSSs that simulate their environmental transport from deposition to mobility in soil and estimate the resulting radiation doses. (3) Extension/update of existing DSSs in the area of atmospheric dispersion on the basis of recent experiences, technological advances and users' requirements. In conclusion, the new features that have been developed regarding atmospheric dispersion and consequence modelling in the frame of PREPARE project, have substantially increased the capabilities of operational DSSs, and have also revealed directions for future research and development
A gamma radiation dose calculation method for use with Lagrangian puff atmospheric dispersion models used in real-time emergency response systems
No full textMethod of Source Identification Following an Accidental Release at an Unknown Location Using a Lagrangian Atmospheric Dispersion Model
No full textA computationally efficient source inversion algorithm was developed and applied with the Lagrangian atmospheric dispersion model DIPCOT. In the process of source location estimation by minimizing a correlation-based cost function, the algorithm uses only the values of the time-integrated concentrations at the monitoring stations instead of all of the individual measurements in the full concentration-time series, resulting in a significant reduction in the number of integrations of the backward transport equations. Following the source location estimation the release start time, duration and emission rate are assessed. The developed algorithm was verified for the conditions of the ETEX-I (European Tracer Experiment—1st release). Using time-integrated measurements from all available stations, the distance between the estimated and true source location was 108 km. The estimated start time of the release was only about 1 h different from the true value, within the possible accuracy of estimate of this parameter. The estimated release duration was 21 h (the true value was 12 h). The estimated release rate was 4.28 g/s (the true value was 7.95 g/s). The estimated released mass almost perfectly fitted the true released mass (323.6 vs. 343.4 kg). It thus could be concluded that the developed algorithm is suitable for further integration in real-time decision support systems
MODELLING OF ATMOSPHERIC FLOW AND DISPERSION IN THE WAKE OF A CYLINDRICAL OBSTACLE
No full textThis paper presents computational simulations of atmospheric dispersion experiments conducted around isolated
obstacles in the field. The computational tool used for the simulations was the code ADREA-HF, which was especially developed
for the simulation of the dispersion of positively or negatively buoyant gases in complicated geometries. The field experiments
simulated involve a single cylindrical obstacle normal to the mean wind direction and two upwind sources of ammonia and propane,
with the ammonia source located at different lateral positions (Mavroidis et al., 2003). Concentrations and concentration fluctuations
for both gases were calculated by the model and compared with the experimental results to evaluate the model performance.
Specific characteristics of dispersion were investigated using the computational tool. Comparisons of experimental and model
results with the case of dispersion around an isolated cubical obstacle are also presented and discussed
Source Term Estimation for Puff Releases Using Machine Learning: A Case Study
No full textReliable source term prediction for hazardous pollutant puffs in urban microenvironments is challenging, especially for risk management under strict time constraints. Puff movement is highly stochastic due to atmospheric turbulence, intensified by complex urban canopies. This complexity, combined with time limitations, makes advanced computational modeling impractical. A more efficient approach is leveraging past and present data using Machine Learning (ML) techniques. This study proposes an ML-based method, enriched with simplified physical modeling, for source term estimation of unforeseen hazardous air releases in monitored urban areas. The Random Forest Regression, commonly used in meteorology and air quality studies, has been selected. A novel variable selection method is introduced, including the following: (a) a model-derived Exposure Burden Index (EBI) reflecting plume–morphology interactions; (b) a plume travel time indicator; (c) the standard deviation of input variables capturing stochastic behavior; and (d) the total dosage-to-mass released ratio at sensor locations as the target variable. The case study examines JU2003 field experiments involving SF6 puffs released at street level in Oklahoma City’s urban core, a challenging scenario due to the limited number of sensors and historical data. Results demonstrate the approach’s effectiveness, offering a promising, realistic alternative to traditional computationally intensive methods
Evaluation of probability distributions for concentration fluctuations in a building array
No full textImprovement of source and wind field input of atmospheric dispersion model by assimilation of concentration measurements: Method and applications in idealized settings
No full textAbstractThe problem of correcting the pollutant source emission rate and the wind velocity field inputs in a puff atmospheric dispersion model by data assimilation of concentration measurements has been considered. Variational approach to data assimilation has been used, in which the specified cost function is minimized with respect to source strength and/or wind field. The analyzed wind field satisfied the constraints derived from the conditions of mass conservation and linearized flow equations for perturbations from the first guess wind field. ‘Identical twin’ numerical experiments have been performed for the validation of the method. The first guess estimation errors of source emission rate and wind field were set to a factor of up to 10 and up to 6m/s respectively. The calculations results showed that in most studied cases an improvement of vector wind difference (VWD) error by about 0.7–1m/s could be achieved. The resulting normalized mean square error (NMSE) of concentration field was also reduced significantly
New functionalities developed in the NERIS-TP project regarding meteorological data used by Decision Support Systems
No full textIn this paper a description is given of software tools that have been developed during
the NERIS-TP project which provide the capability to users of Decision Support Systems
(DSSs), such as JRODOS, to calculate their own prognostic meteorological data with the
desired spatial and temporal resolution. These tools increase the flexibility of applying
the DSSs for any location on the Earth. This is achieved by downloading freely available
global meteorological data and downscaling them using the prognostic meteorological model
WRF. The results of WRF and/or global data are uploaded to the DSS to calculate the
atmospheric dispersion. The above software tools operate in an automated way, in
conjunction with the DSS. In addition, data assimilation methodologies have been
integrated into the Meteorological Pre-Processor of JRODOS, to correct previously
calculated prognostic data on the basis of locally measured meteorological data. These
data assimilation methods were successfully tested and their results increase the accuracy
of the prognoses of dispersion models
- …
