Boruta-grid-search least square support vector machine for NO2 pollution prediction using big data analytics and IoT emission sensors

PurposeThis paper seeks to assess the performance levels of BA-GS-LSSVM compared to popular standalone algorithms used to build NO2 prediction models. The purpose of this paper is to pre-process a relatively large data of NO2 from Internet of Thing (IoT) sensors with time-corresponding weather and traffic data and to use the data to develop NO2 prediction models using BA-GS-LSSVM and popular standalone algorithms to allow for a fair comparison.Design/methodology/approachThis research installed and used data from 14 IoT emission sensors to develop machine learning predictive models for NO2 pollution concentration. The authors used big data analytics infrastructure to retrieve the large volume of data collected in tens of seconds for over 5 months. Weather data from the UK meteorology department and traffic data from the department for transport were collected and merged for the corresponding time and location where the pollution sensors exist.FindingsThe results show that the hybrid BA-GS-LSSVM outperforms all other standalone machine learning predictive Model for NO2 pollution.Practical implicationsThis paper's hybrid model provides a basis for giving an informed decision on the NO2 pollutant avoidance system.Originality/valueThis research installed and used data from 14 IoT emission sensors to develop machine learning predictive models for NO2 pollution concentration.

Air Pollution Observations in Selected Locations in Poland during the Lockdown Related to COVID-19

The coronavirus disease (COVID-19) has caused huge changes in people’s daily habits and had a significant impact on the economy. The lockdowns significantly reduced road traffic and meant that many people worked remotely. Therefore, the question arose as to how the reduced road traffic and stays of residents at home affected the degree of pollution and the structure of major air pollutants. To answer this question, the article presents an analysis of changes in typical air pollutants (PM10, PM2.5, NO2) in the five largest Polish cities and one of the voivodships. The data from the Polish State Environmental Monitoring were used for the analysis. The analysis showed that the period of the first lockdown in Poland (April 2020), despite the reduced road traffic, resulted in a significant increase in PM10 emissions (9–91% during working days and an average of 30% on Saturdays and Sundays), a slight increase in PM2.5 emissions (on average from 2% to 11% for all analyzed locations), and a reduction in NO2 emissions (on average from 6% to 11% for all analyzed locations) compared to the period before the lockdown. However, the changes were not homogeneous—in Łódź and Warsaw, in most cases, an increase in all analyzed pollutants was observed, and the greatest decrease in pollution took place in Małopolska voivodship (including Kraków). Comparing the data from April 2020 to the data from April 2019, the overall difference in the PMs concentrations was small, although there are places where there has been a significant decrease (Wrocław, Poznań), and there were also places where the concentration increased (Warsaw, Łódź, Małopolska). In the case of nitrogen dioxide, pollution concentration decreased in most locations. The only exception was the background stations in Warsaw, where the increase was 27%.

Multi-scale analysis of turbulence data from AIRTEC-CM urban field campaigns in Madrid

<p>Several urban field campaigns have been carried out in the city of Madrid (Spain) during the years 2020 and 2021 in the frame of the AIRTEC-CM <sup>(*)</sup> research project (Urban Air Quality and Climate Change Integral Assessment). The analysis of the relation between the turbulence measured close to the surface and pollution concentration (e.g., particle matter of different sizes, NO<sub>x</sub>, etc.) is a key aspect to achieve the different objectives of the project. In this context, we present some preliminary analyses of the turbulence data measured from sonic anemometers located at different emplacements. We focus on the turbulence differences among two instruments nearby located but at different heights above the street level: 1) at the top of a 22 m-height building, and 2) at the top of a shorter building of 2.5 m of height. Typical turbulent parameters (turbulent kinetic energy (TKE), friction velocity (u*) and sensible heat flux (SH)) are analysed for both sonic anemometers and their differences are statistically compared. An investigation of the main temporal scales involved in the atmospheric diffusion is also performed using the Multi-Resolution Flux Decomposition technique (MRFD), applied over a relatively long period that includes different atmospheric conditions in February 2020. The information obtained from this analysis will be related to the pollution concentration measured in the city, trying to determine the importance of the near-surface turbulence (and the corresponding scales of the main eddies found) in the pollutant’s levels.</p><p> </p><p>(*) AIRTEC-CM research project (S2018/EMT-4329) is funded by The Regional Government of Madrid (Spain).</p>

Managing the Balance of Consumption and Reproduction of Marine Resources in the “Coast – Sea” Ecological-Economic System

Purpose. The work is aimed at constructing an adaptive model of the ecological-economic system of the sea coastal zone intended to control the volume of consumption of marine assimilation and biological resources by the coastal socio-economic system. Methods and Results. The model based on tracking the integral balance of the rates of marine environment pollution and its purification due to the natural physical and biogeochemical processes and to the environmental actions is proposed. The average multi-year values of the ecosystem variables are used in the applied concept of stationary state of the marine ecosystem. To describe the conditions forming the polluted area in the coastal marine environment, proposed is the algorithm taking into account the ratio between the rate of growth of concentration of the pollutants inflowing from the coastal sources and the rate of their destruction resulting from natural purification of marine environment. Coastal pollution runoff was managed by the penalties imposed on the economic system, which was forced to reduce the generalized product output simultaneously increasing its prime cost. In the marine ecosystem model, the main lower links of the food chain were used: concentrations of phytoplankton, zooplankton and bacterioplankton. Their weighted sum constituted the marine biodiversity index, which was taken as a concentration of bio-resource consumed by the coastal economic system. It was believed that there was an inverse relationship between the biodiversity index and the pollution concentration. Therefore, the maximum permissible pollution concentration was assessed based on the permissible values of the biodiversity index. Along with control of the economic efficiency of marine resource consumption, the model provided for the biodiversity index monitoring so that it did not fall below its permissible vales. In this case, the management agent included in the model, limited production. The model equations are constructed by the method of adaptive balance of causes, which preserved the material balances resulting from the interaction between the environmental and economic processes. The modular equations of the method permitted to use the known stationary state of the ecosystem to obtain the estimates of the influence coefficients. It is shown that the forecasted scenarios provide the compromises between the efficiency of the coastal economic system and the ability of marine ecosystem to restore the consumed assimilation and biological resources. Conclusions. The proposed model makes it possible to forecast the scenarios of the ecological and economic processes that provide a balance of consumption and reproduction of marine resources.