Air pollution data classification by SOM Neural Network

Over the last ten years, Salamanca has been considered among the most polluted cities in México. This paper presents a Self-Organizing Maps (SOM) Neural Network application to classify pollution data and automatize the air pollution level determination for Sulphur Dioxide (SO2) in Salamanca. Meteorological parameters are well known to be important factors contributing to air quality estimation and prediction. In order to observe the behavior and clarify the influence of wind parameters on the SO2 concentrations a SOM Neural Network have been implemented along a year. The main advantages of the SOM is that it allows to integrate data from different sensors and provide readily interpretation results. Especially, it is powerful mapping and classification tool, which others information in an easier way and facilitates the task of establishing an order of priority between the distinguished groups of concentrations depending on their need for further research or remediation actions in subsequent management steps. The results show a significative correlation between pollutant concentrations and some environmental variables.

Caracterización experimental del centelleo troposférico en banda Ka

El presente Proyecto Fin de Grado tiene como objetivo el estudio y caracterización del centelleo troposférico en ausencia de lluvia en la banda Ka de un enlace Tierra-satélite. Para ello se dispondrá de un equipo receptor situado en la Escuela Técnica Superior de Ingenieros de Telecomunicación. Los datos son emitidos desde el satélite EutelSat Hot Bird 13A a una frecuencia de 19,7 GHz. La primera parte del proyecto comienza con las bases teóricas de los distintos fenómenos que afectan a la propagación de un enlace satélite, mencionando los modelos de predicción más importantes. Se ha dado más importancia al apartado perteneciente al centelleo troposférico por ser el tema tratado en este proyecto. El estudio cuenta con datos del satélite durante 7 años comprendidos entre julio de 2006 a junio de 2013. Después del filtrado y el resto del tratamiento adecuado de los datos se han obtenido distintas distribuciones estadísticas que están relacionadas con el centelleo como la varianza. Más tarde se ha comparado la varianza experimental con parámetros meteorológicos obtenidos desde distintas bases de datos. El objetivo de esto ha sido discernir cuál de estos factores afecta en mayor medida a la intensidad de centelleo. Para ello se ha realizado la correlación entre la varianza y varios parámetros meteorológicos: temperatura, humedad relativa, humedad absoluta, índice de refracción húmedo, presión… Además se han realizado medidas de nubosidad en los que se ha clasificado las muestras dependiendo del tipo de nube presente en el cielo. A continuación se ha calculado la varianza mensual media y distribuciones acumuladas de ciertos modelos de predicción de centelleo, comparándolos gráficamente con las curvas experimentales. Estos modelos usan parámetros medidos en superficie por lo que se utilizarán algunos de los parámetros analizados en el capítulo anterior. Por último se expondrán las conclusiones sacadas a lo largo de la realización del proyecto y las posibles líneas de investigación futuras. ABSTRACT. The present Project has as the principal aim the study and characterization of tropospheric scintillation in lack of rain in the band Ka of an Earth-satellite link. It is provided for a receptor equipment located in the ETSIT. The data are broadcasted form the Eutelsat Hot Bird 13A satellite at the frecuency of 19,7 GHz. The beginning of the project starts with the theorical basis of the different phenomenons that affects to the propagation of a satellite link, naming the most important predictions models. The chapter referred to the scintillation has had more importance due to be the main topic in this project. The study deals with satellite data during 7 years between July 2006 to June 2013. After the filter and others treatments of the data, it has been getting different statistics distributions related to scintillation like variance. Later, the experimental variance has been compared with meteorological parameters obtained from different datasets. The purpose has been to decide which factor affects in a greater way to the scintillation intensity. For that it has been doing the correlation between variance and meteorological parameters: temperature, relative humidity, absolute humidity, air refractivity due to water vapour, pressure… Moreover, it has been doing cloudiness measurements in which the samples have been classified in order to the kind of cloud shown in the sky at that moment. Then it has been calculated the monthly averaged variance and the prediction model for cumulative distributions which has been compared with the experimental results. That models uses surface data that they will be uses some meteorological parameters analyzed in previous chapters. Finally it will be shown the conclusions obtained along the realization of the project and the possible ways of future research.

Study of PM10 and PM2.5 levels in three European cities: Analysis of intra and inter urban variations

In the present paper, 1-year PM10 and PM 2.5 data from roadside and urban background monitoring stations in Athens (Greece), Madrid (Spain) and London (UK) are analysed in relation to other air pollutants (NO,NO2,NOx,CO,O3 and SO2)and several meteorological parameters (wind velocity, temperature, relative humidity, precipitation, solar radiation and atmospheric pressure), in order to investigate the sources and factors affecting particulate pollution in large European cities. Principal component and regression analyses are therefore used to quantify the contribution of both combustion and non-combustion sources to the PM10 and PM 2.5 levels observed. The analysis reveals that the EU legislated PM 10 and PM2.5 limit values are frequently breached, forming a potential public health hazard in the areas studied. The seasonal variability patterns of particulates varies among cities and sites, with Athens and Madrid presenting higher PM10 concentrations during the warm period and suggesting the larger relative contribution of secondary and natural particles during hot and dry days. It is estimated that the contribution of non-combustion sources varies substantially among cities, sites and seasons and ranges between 38-67% and 40-62% in London, 26-50% and 20-62% in Athens, and 31-58% and 33-68% in Madrid, for both PM10 and PM 2.5. Higher contributions from non-combustion sources are found at urban background sites in all three cities, whereas in the traffic sites the seasonal differences are smaller. In addition, the non-combustion fraction of both particle metrics is higher during the warm season at all sites. On the whole, the analysis provides evidence of the substantial impact of non-combustion sources on local air quality in all three cities. While vehicular exhaust emissions carry a large part of the risk posed on human health by particle exposure, it is most likely that mitigation measures designed for their reduction will have a major effect only at traffic sites and additional measures will be necessary for the control of background levels. However, efforts in mitigation strategies should always focus on optimal health effects.