Spatio-temporal circular models with non-separable covariance structure

@article{Mastrantonio2016SpatiotemporalCM,
  title={Spatio-temporal circular models with non-separable covariance structure},
  author={Gianluca Mastrantonio and Giovanna Jona Lasinio and Alan E. Gelfand},
  journal={TEST},
  year={2016},
  volume={25},
  pages={331-350}
}
Circular data arise in many areas of application. Recently, there has been interest in looking at circular data collected separately over time and over space. Here, we extend some of this work to the spatio-temporal setting, introducing space–time dependence. We accommodate covariates, implement full kriging and forecasting, and also allow for a nugget which can be time dependent. We work within a Bayesian framework, introducing suitable latent variables to facilitate Markov chain Monte Carlo… 
CircSpaceTime: an R package for spatial and spatio-temporal modelling of circular data
TLDR
CircSpaceTime is the only R package, currently available, that implements Bayesian models for spatial and spatio-temporal interpolation of circular data, and the proposed models are cast as hierarchical, with fitting and inference within a Bayesian framework.
CircSpaceTime: an R package for spatial and spatio-temporal modeling of Circular data
TLDR
The proposed models are cast as hierarchical, with fitting and inference within a Bayesian framework, and the implementation of the proposed models is considerably improved on the simple routines adopted in the research papers.
Likelihood-based inference for multivariate space-time wrapped-Gaussian fields
TLDR
A wrapped-Gaussian field is proposed to model directions in a multivariate spatial or spatio-temporal context and the approach outperforms the analytical and computational limitations of full likelihood, because it works with the marginal bivariate distributions of the random field.
The wrapped skew Gaussian process for analyzing spatio-temporal data
TLDR
A novel wrapped skew Gaussian process is proposed which enables more flexible marginal distributions than the symmetric ones arising under the wrappedGaussian process and it allows straightforward interpretation of parameters.
The Joint Projected and Skew Normal
TLDR
A new multivariate circular linear distribution suitable for modeling direction and speed in (multiple) animal movement data is introduced and its predictive performance is compared to those given by hidden Markov models built on all the combinations of von Mises, wrapped Cauchy, gamma, gamma and Weibull distributions.
A Bayesian hidden Markov model for telemetry data
TLDR
A new multivariate circular linear distribution suitable for modeling direction and speed in (multiple) animal movement data is introduced and its predictive performance is compared to those given by hidden Markov models built on all the combinations of von Mises, wrapped Cauchy, gamma, gamma and Weibull distributions.
A smoothing spline model for multimodal and skewed circular responses: Applications in meteorology and oceanography
The analysis of circular data is the main subject in many disciplines, such as meteorology and oceanography. In this article, we introduce a new multimodal skew‐circular model as an extension of the
Clustering of longitudinal interval-valued data via mixture distribution under covariance separability
TLDR
The use of a separable covariance matrix introduces several advantages in M-clustering, which include fewer samples required for a valid procedure and finding the correct number of clusters more accurately compared to other commonly assumed covariance matrices.
Random fields on the hypertorus: Covariance modeling and applications
TLDR
This article gives a comprehensive theoretical framework to the modeling, inference, and applications of Gaussian random fields using what is term the hypertorus as an index set and develops various families of covariance functions over thehypertorus.
The joint projected and skew normal ; a distribution for poly-cylindrical data
TLDR
A multivariate circular-linear distribution obtained by combining the projected and the skew normal is introduced, enabling a Markov chain Monte Carlo algorithm relying only on Gibbs steps, where the updates of the parameters are done as if the authors were working with a multivariate normal likelihood.
...
...

References

SHOWING 1-10 OF 28 REFERENCES
Modeling Space and Space-Time Directional Data Using Projected Gaussian Processes
TLDR
The contribution is to develop a fully model-based approach to capture structured spatial dependence for modeling directional data at different spatial locations, and builds a projected Gaussian spatial process, induced from an inline bivariate Gaussia spatial process.
Bayesian hidden Markov modelling using circular‐linear general projected normal distribution
We introduce a multivariate hidden Markov model to jointly cluster time‐series observations with different support, that is, circular and linear. Relying on the general projected normal distribution,
Directional data analysis under the general projected normal distribution.
Spatial analysis of wave direction data using wrapped Gaussian processes
TLDR
A model-based approach to handle periodic data in the case of measurements taken at spatial locations, anticipating structured dependence between these measurements, and formulates a wrapped Gaussian spatial process model for this setting.
Joint spatio-temporal analysis of a linear and a directional variable: space-time modeling of wave heights and wave directions in the Adriatic Sea
It is valuable to have a better understanding of factors that influence sea motion and to provide more accurate forecasts. In particular, we are motivated by data on wave heights and outgoing wave
Nonseparable, Stationary Covariance Functions for Space–Time Data
Geostatistical approaches to spatiotemporal prediction in environmental science, climatology, meteorology, and related fields rely on appropriate covariance models. This article proposes general
Inference for circular distributions and processes
  • S. Coles
  • Computer Science, Mathematics
    Stat. Comput.
  • 1998
TLDR
The power and flexibility of Markov chain Monte Carlo methods to fit such classes of models to circular data and applications are given to multivariate and time series data of wind directions.
A Multivariate Hidden Markov Model for the Identification of Sea Regimes from Incomplete Skewed and Circular Time Series
The identification of sea regimes from environmental multivariate times series is complicated by the mixed linear–circular support of the data, by the occurrence of missing values, by the skewness of
Bayesian Analysis of Circular Data Using Wrapped Distributions
Circular data arise in a number of different areas such as geological, meteorological, biological and industrial sciences. Standard statistical techniques can not be used to model circular data due
A Markov process for circular data
TLDR
A discrete time Markov process which takes values on the unit circle is proposed, which is compared with some existing Markov processes for circular data and an application of the model to wind direction data is provided.
...
...