Triple seasonal methods for short-term electricity demand forecasting

Online short-term load forecasting is needed for the real-time scheduling of electricity generation. Univariate methods have been developed that model the intraweek and intraday seasonal cycles in intraday load data. Three such methods, shown to be competitive in recent empirical studies, are double seasonal ARMA, an adaptation of Holt–Winters exponential smoothing for double seasonality, and another, recently proposed, exponential smoothing method. In multiple years of load data, in addition to intraday and intraweek cycles, an intrayear seasonal cycle is also apparent. We extend the three double seasonal methods in order to accommodate the intrayear seasonal cycle. Using six years of British and French data, we show that for prediction up to a day-ahead the triple seasonal methods outperform the double seasonal methods, and also a univariate neural network approach. Further improvement in accuracy is produced by using a combination of the forecasts from two of the triple seasonal methods.     

Rainfall frequency and seasonality identification through artificial neural networks

The artificial neural network technique is experimented to cope with the study of the sub-annual seasonal non-stationarity of the rainfall process. The homogeneity of the climatic signals inside each of the natural 12 monthly classes is analyzed, adopting a multilayer feed-forward network with error back-propagation. The possibility of identifying monthly based seasons from only daily rainfall data is found to be quite limited. The coupling of rainfall and temperature statistics is instead confirmed to be a fundamental climatic indicator. Contrary to what is commonly expected, the season uncertainty appears higher in summer and in winter than in spring or autumn. The hypothesis of defining any monthly based pluviometric regime is however demonstrated to be generally difficult to sustain, revealing the necessity of adopting an unsupervised criterion to identify any seasonal filter of the rainfall process.

---

Sommario Viene sperimentata la tecnica delle reti neurali artificiali per affrontare lo studio della non-stazionarità stagionale del processo di precipitazione. Viene analizzata l'omogeneità dei segnali climatici all'interno di ciascuna delle 12 naturali classi mensili, adottando un rete di tipo multistrato feed-forward con retro-propagazione dell'errore. La possibilità di indentificare le stagioni su base mensile con i soli dati di precipitazione si dimostra essere piuttosto limitata. Viceversa, l'accoppiamento fra le statistiche di pioggia e temperatura si rivela essere un fondamentale indicatore climatico. Al contrario di quanto comunemente reputato, l'incertezza nella stagionalità appare essere più alta in estate ed in inverno che in primavera e in autunno. L'ipotesi sulla definizione su base mensile dei regimi pluviometrici appare comunque difficile da sostenere, rivelando la necessità di adottare un criterio senza supervisione per l'identificazione del filtro stagionale relativo al processo di precipitazione.

     

On Modelling and Pricing Rainfall Derivatives with Seasonality

Abstract

We are interested in pricing rainfall options written on precipitation at specific locations. We assume the existence of a tradeable financial instrument in the market whose price process is affected by the quantity of rainfall. We then construct a suitable ‘Markovian gamma’ model for the rainfall process which accounts for the seasonal change of precipitation and show how maximum likelihood estimators can be obtained for its parameters.

We derive optimal strategies for exponential utility from terminal wealth and determine the utility indifference price of the claim. The method is illustrated with actual measured data on rainfall from a location in Kenya and spot prices of Kenyan electricity companies.