Daily rainfall forecasting using an ensemble technique based on singular spectrum analysis

Studer and Masulli (1995), Masulli, Parenti, and Studer (1999), and Masulli, Cicione, and Studer (2000) proposed a constructive methodology for temporal data learning supported by results and prescriptions related to the Takens-Mane theorem and using the singular spectrum analysis in order to reduce the effects of the possible discontinuity of the signal. In this paper we present some new results concerning the application of this approach to the forecasting of the individual rainfall intensities series collected by 135 stations distributed in the Tiber basin.