Neural Networks for Environmental Problems: Data Quality Control and Air Pollution Nowcasting

This work illustrates the use and some related results of Artificial Neural Networks (ANNs) for data
quality control of environmental time series and for reconstruction of missing data. ANNs are applied
to the following problems: i) short and medium-term predicting of air pollutant concentrations in urban
areas, ii) interpolating and extrapolating daily maximum temperature, iii) replacing time distribution
with spatial distributed information (pollutant concentrations at different measuring sites). Observed
versus predicted data are compared to test the efficacy of ANNs in simulating environmental processes.
Results confirm ANNs as an improvement of classical models and show the utility of ANNs for
restoration of time series..