Forecasting of Chaotic Cloud Absorption Time Series for Meteorological and Plume Dispersion Modeling

Abstract

A nonlinear forecasting method based on the reconstruction of a chaotic strange attractor from about 1.5 years of cloud absorption data obtained from half-hourly Meteosat infrared images was used to predict the behavior of the time series 24 h in advance. The forecast values are then used by a meteorological model for daily prediction of plume transport from the As Pontes 1400-MW power plant in northwestern Spain. Results from the meteorological model, using the cloud absorption predictions, are compared with measurements obtained from meteorological towers and a Remtech PA-3 sodar. The effects of cloud absorption on SO2 ground-level concentration forecasts are analyzed for two different days.