Daniel Dunea

Using wavelet–feedforward neural networks to improve air pollution forecasting in urban environments

The paper presents the screening of various feedforward neural networks (FANN) and wavelet–feedforward neural networks (WFANN) applied to time series of ground-level ozone (O3), nitrogen dioxide (NO2), and particulate matter (PM10 and PM2.5 fractions) recorded at four monitoring stations located in various urban areas of Romania, to identify common configurations with optimal generalization performance. Two distinct model runs were performed as follows: data processing using hourly-recorded time series of airborne pollutants during cold months (O3, NO2, and PM10), when residential heating increases the local emissions, and data processing using 24-h daily averaged concentrations (PM2.5) recorded between 2009 and 2012. Dataset variability was assessed using statistical analysis. Time series were passed through various FANNs. Each time series was decomposed in four time-scale components using three-level wavelets, which have been passed also through FANN, and recomposed into a single time series. The agreement between observed and modelled output was evaluated based on the statistical significance (r coefficient and correlation between errors and data). Daubechies db3 wavelet–Rprop FANN (6-4-1) utilization gave positive results for O3 time series optimizing the exclusive use of the FANN for hourly-recorded time series. NO2 was difficult to model due to time series specificity, but wavelet integration improved FANN performances. Daubechies db3 wavelet did not improve the FANN outputs for PM10 time series. Both models (FANN/WFANN) overestimated PM2.5 forecasted values in the last quarter of time series. A potential improvement of the forecasted values could be the integration of a smoothing algorithm to adjust the PM2.5 model outputs.

Fine Particulate Matter in Urban Environments: A Trigger of Respiratory Symptoms in Sensitive Children

The overall objective of this research was to study children’s respiratory illness levels in Targoviste (Romania) in relationship to the outdoor concentrations of airborne particulate matter with an aerodynamic diameter below 2.5 µm (PM2.5). We monitored and analysed the PM2.5 concentrations according to a complex experimental protocol. The health trial was conducted over three months (October–December 2015) and required the active cooperation of the children’s parents to monitor carefully the respiratory symptoms of the child, i.e., coughing, rhinorrhoea, wheezing, and fever, as well as their outdoor program. We selected the most sensitive children (n = 25; age: 2–10 years) with perturbed respiratory health, i.e., wheezing, asthma, and associated symptoms. The estimated average PM2.5 doses were 0.8–14.5 µg·day−1 for weekdays, and 0.4–6.6 µg·day−1 for the weekend. The frequency and duration of the symptoms decreased with increasing age. The 4- to 5-year old children recorded the longest duration of symptoms, except for rhinorrhoea, which suggested that this age interval is the most vulnerable to exogenous trigger agents (p < 0.01) compared to the other age groups. PM2.5 air pollution was found to have a direct positive correlation with the number of wheezing episodes (r = 0.87; p < 0.01) in November 2015. Monitoring of wheezing occurrences in the absence of fever can provide a reliable assessment of the air pollution effect on the exacerbation of asthma and respiratory disorders in sensitive children.

Investigation and Selection of Remediation Technologies for Petroleum-Contaminated Soils Using a Decision Support System

We have developed a decision support system that prompts the user for site and contaminant characteristics from petroleum extraction and transportation processes and filters out those technologies that are most adequate for site remediation from the categories of remedial methods in its database. The computerized rule-based structure facilitates the evaluation and selection of the most applicable and feasible treatment technology for petroleum-contaminated soil on a site polluted from pipeline deterioration, which this paper presents as a case study. The technology-screening feature is a useful tool in the early evaluation phase of the soil remediation process and can assist specialists in the simulation of various scenarios. The system has accurately estimated pollutant migration and extent for preliminary soil investigations and provided recommendations in accordance with the current remedial actions taken at these petroleum-contaminated sites. The decision elaboration is close to the situations often encountered in real life.

A Cyberinfrastructure for Air Quality Monitoring and Early Warnings to Protect Children with Respiratory Disorders

Air pollution and health are closely connected. An efficient management of air quality involves the collecting and analysis of a wide variety of data types that make cyber infrastructures especially important for this scientific field. In the new information-driven world, a cyber infrastructure allows to bring people together, high performance computational platforms, data integration techniques, visualization, and analysis tools, with the aim to make the research more effective and efficient. Currently, there is an important requirement for the development of information networks that are specific to air quality management, promoting the exchange and integration of data and applications. The paper presents the developing of a cyber infrastructure for air quality management to protect childrens health in Romania developed by ROKIDAIR research project. This infrastructure combines already existing air quality monitoring systems with new modes of data processing and displaying to be accessed by the concerned parents of children.