Application of Bayesian approach and probability theory for calculating territorial risk and zoning on urban area

• Published in: Journal of physics. Conference series 2020-12, Vol.1703 (1), p.12045
• Main Authors: Tunakova, Y A, Shagidullin, A R, Valiev, V S, Galimova, A R
• Format: Article
• Language: English
• Subjects: Algorithms; Arrays; Bayesian analysis; Cartography; Model accuracy; Model testing; Nonuniformity; Physics; Probabilistic methods; Probabilistic models; Probability; Probability theory; Risk; Statistical analysis; Urban areas; Zoning
• ISSN: 1742-6588; 1742-6596
• DOI: 10.1088/1742-6596/1703/1/012045

The article describes a new algorithm for assessing territorial risk. The risk in the algorithm is calculated as the ratio of the prior and posterior probabilities of an unfavorable (risk) event. An event is an excess of the threshold content of toxicants in a certain area of space, taking into account how often this event occurred in all other areas of space. This can be used for zoning an urbanized area. We used a probabilistic safety assessment (PSA) apparatus based on hazard modelling and a scenario-based approach. Such an apparatus does not analyse certain integral characteristics of the process, but all possible states in a given time section at the same time, forming a probabilistic model. The probabilistic approach ensures the accuracy and realism of assessments of the state of the environment of an urbanized area that are unattainable for deterministic methods and, in addition, allows you to take into account the cumulative effect of multiple sources of uncertainty. Using the probabilistic method and assessing the level of risks allows you to broadcast complex scientific information in an accessible format to people making environmental decisions. Risk is not an absolute but a relative category. It is heterogeneous in both time and space. Therefore, it can be used when zoning the environment, especially since the parameters that are used to calculate risk very often form the same data arrays necessary for classifying and zoning spatial non-uniformity of the system in general. After analyzing various approaches, we came to the conclusion that for a given time period the probability of an adverse event will be calculated more accurately if we calculate not using the classical formula (the number of cases per time interval), but in the form of the ratio of the posterior and prior probabilities. This takes into account the risk in a certain area of space and correlates with the frequency of the same event in all other areas. To assess the likelihood of polymetallic pollution of environmental objects in an urbanized area with its subsequent zoning, we considered the variability of the metal content in environmental objects. Based on the probabilities that metal concentrations exceeded their threshold values, the territorial relative risk was calculated and zoning of the urbanized territory was carried out. Testing of the considered approaches was carried out in the territory of Kazan on the basis of an array of data of experimental measurements on the