Severe air pollution episodes in Florence: data analysis and forecasting operational method

摘要

An operational method has been developed to forecast the maximum hourly concentration of NO_2 in Florence urban area during the current day. The situations that can lead to high concentrations of NO_2 have been identified and described by means of a statistical analysis of the environmental conditions measured. The forecasting method has been developed through the following tasks: 1. preliminary study to select the most significant environmental parameters and synoptic conditions; 2. characterisation of each day using the data measured till 10 am; 3. selection of the best algorithm to compare the current day with the past events and to individuate the days alike; 4. use of statistical methods to calculate the probability to exceed the NO_2 warning limits during the next hours.