Improvement of surface water quality variables modelling that incorporates a hydro-meteorological factor: a state-space approach

Abstract

In this work it is constructed a hydro-meteorological factor to improve the adjustment of statistical time series models, such as state space models, of water quality variables by observing hydrological series (recorded in time and space) in a River basin. The hydro-meteorological factor is incorporated as a covariate in multivariate state space models fitted to homogeneous groups of monitoring sites. Additionally, in the modelling process it is considered a latent variable that allows incorporating a structural component, such as seasonality, in a dynamic way.