Environmetrics has just published a new special issue on Joint Outcome Modeling in Environmetrics. The special issue, guest edited by Charmaine B. Dean and Elizabeth M. Renouf, examines the benefits of using joint outcome modeling in environmental contexts. 

A portion of the special issue’s editorial, written by Dean and Renouf, and a list of papers included in the special issue are included below. 

A commonly used method in epidemiology links multiple outcomes such as survival and longitudinal data in a so-called joint outcome model. This analytical approach allows the simultaneous analysis of different types of correlated outcomes on an individual in a single model. Similarly, biologists and ecologists are often interested in the relationship between multiple types of outcomes such as, for example, species survival and breeding success, incorporating many types of correlations beyond individual-level associations. The use of joint models in ecology and environmetrics has pioneered new developments of more flexible, complex and creative modeling methods. In particular, researchers faced with different data types to be simultaneously modeled have eagerly embraced the joint modeling framework, exploring such new approaches. Beyond accounting for potential correlations in outcomes explicitly, advantages include the ability to quantify such correlations, as well as potential efficiencies in analysis. The approach also allows better understanding of related processes and how they influence each other. The aim of this special issue is to identify advantages of using the joint modeling framework in the environmental context.

Our aim in this special issue is to highlight some interesting work and motivate future applications of the joint model in environmetrics. In this issue, the applications span a wide range of topics. In climate science, Konzen et al. (2021) provide a framework for inference on storm peak significant wave heights at multiple locations in the North Sea. Xi et al. (2021) present an analysis of extreme wildland fire behaviour, modeling the duration and size of fires as joint outcomes. North et al. (2020) employ a joint statistical framework for an analysis of the minimum and maximum temperature cycles for the years 1996 to 2018 at almost one hundred thousand locations. Zhang et al. (2021) utilize a joint modeling framework for analysis of very large spatial datasets with over a million locations, illustrated by an application to a study of vegetation structure and its spatiotemporal variation. Finally, Niku et al. (2021) employ a latent variable model in an application of a joint model for species abundance, and also provide a validation of type 1 errors for interaction tests. Environmetrics data offer a wealth of opportunities for the application of a joint modeling framework. The range of the articles included in this issue, considering both the research topics and the statistical approaches, demonstrate the broad usefulness of the joint modeling framework. The special issue also motivates and identifies significant opportunities to address important gaps in joint modeling approaches to environmental research.

High-dimensional multivariate geostatistics: A Bayesian matrix-normal approach by Lu Zhang, Sudipto Banerjee and Andrew O. Finley 

Analyzing environmental-trait interactions in ecological communities with fourth-corner latent variable models by Jenni Niku, Francis K. C. Hui, Sara Taskinen and David I. Warton (Open Access)

Modeling nonstationary extremes of storm severity: Comparing parametric and semiparametric inference by Evandro Konzen, Cláudia Neves and Philip Jonathan (Open Access)

On the spatial and temporal shift in the archetypal seasonal temperature cycle as driven by annual and semi-annual harmonics by Joshua S. North, Erin M. Schliep and Christopher K. Wikle

Modeling the duration and size of wildfires using joint mixture models by Dexen D. Z. Xi, Charmaine B. Dean and Stephen W. Taylor