Spherically Embedded Time Series with Unknown Trend and Periodic Components

TL;DR

This paper introduces a geometric framework for modeling nonstationary spherically embedded time series, utilizing a spherical trend-periodicity decomposition and spherical autoregressive models for improved forecasting.

Contribution

It presents a novel nonparametric spherical decomposition method and a trend-periodic spherical autoregressive model, addressing the gap in analyzing nonstationary spherical time series.

Findings

The proposed model achieves superior predictive performance in simulations.

The methodology provides interpretable insights into structural dynamics.

Applications to real data improve forecasting accuracy.

Abstract

Spherically embedded time series are time series with values naturally residing on or can be equivalently mapped to the sphere. Despite their ubiquity in diverse scientific fields, these data frequently exhibit complex non-stationarity driven by latent trend and periodic components. Traditional Euclidean time series methods fail to account for the intrinsic non-Euclidean geometry of the sphere, leaving a critical gap in rigorous methodologies for modelling and forecasting nonstationary spherically embedded time series. To address this methodological gap, we propose a unified geometric framework to analyse nonstationary spherically embedded time series. Central to our approach is a novel nonparametric spherical trend-periodicity decomposition model that uses an optimal-transport-based removal operation to sequentially extract the smooth trend and periodic components while preserving spherical topology. The resulting de-trended and de-seasonalised stationary residuals can be further modelled using a spherical autoregressive model, formalising a novel trend-periodic spherical autoregressive model. Theoretical foundations for the modelling procedure are established on the consistency under temporal dependence. Extensive simulations corroborate these theoretical guarantees and demonstrate the superior finite-sample predictive performance of the trend-periodic spherical autoregressive model. Finally, we validate the practical utility of our methodology through applications to electricity generation compositions and bike trip volume profiles, yielding significantly enhanced forecasting accuracy while providing interpretable insights into the underlying structural dynamics.