Abstract
AbstractMotivated by real‐world data of monthly values of precipitation, minimum, and maximum temperature recorded at 360 monitoring stations covering the Italian territory for 60 years ( months), in this work we propose a change‐point model for multiple multivariate time series, inspired by the hierarchical Dirichlet process. We assume that each station has its change‐point structure and, as main novelties, we allow unknown subsets of the parameters in the data likelihood to stay unchanged before and after a change‐point, that stations possibly share values of the same parameters and that the unknown number of weather regimes is estimated as a random quantity. Owing to the richness of the formalization, our proposal enables us to identify clusters of spatial units for each parameter, evaluate which parameters are more likely to change simultaneously, and distinguish between abrupt changes and smooth ones. The proposed model provides useful benchmarks to focus monitoring programs regarding ecosystem responses. Results are shown for the whole data, and a detailed description is given for three monitoring stations. Evidence of local behaviors includes highlighting differences in the potential vulnerability to climate change of the Mediterranean ecosystems from the Temperate ones and locating change trends distinguishing between continental plains and mountain ranges.
Highlights
Climate elements and regimes, such as temperature, precipitation, their annual cycles and mutual relationships, primarily affect the type and distribution of plants, animals, and soils, as well as their combination in complex ecosystems and ecoregions (Bailey, 2004; Metzger et al, 2013)
The approach we propose is motivated by methods that consider time series as possibly broken down into time regimes composed of adjacent observations (Samé et al, 2011), assuming that observations belonging to the same regime follow a common distribution
The main novel contributions of our proposal to the methodology of CP modeling have to do with: i) the definition of CP as due to an unspecified subset of the parameters of a multivariate time series, ii) the consequent identification of the parameters corresponding to the CPs, iii) the estimate of the unknown number of CPs as a by-product of the model fitting process, iv) the fact that parameter values are possibly shared by different monitoring stations forming clusters of spatial units, and v) the ability to distinguish between abrupt changes and smooth ones
Summary
Climate elements and regimes, such as temperature, precipitation, their annual cycles and mutual relationships, primarily affect the type and distribution of plants, animals, and soils, as well as their combination in complex ecosystems and ecoregions (Bailey, 2004; Metzger et al, 2013). The main novel contributions of our proposal to the methodology of CP modeling have to do with: i) the definition of CP as due to an unspecified subset of the parameters of a multivariate time series, ii) the consequent identification of the parameters (i.e., of the distributional features) corresponding to the CPs, iii) the estimate of the unknown number of CPs as a by-product of the model fitting process, iv) the fact that parameter values are possibly shared by different monitoring stations forming clusters of spatial units, and v) the ability to distinguish between abrupt changes and smooth ones. The Appendix contains details of the MCMC algorithm, an example of how CPs are appropriately identified with artificial data, and the legend of the Italian ecoregion system
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
