Interdisciplinary approaches: towards new statistical methods for phenological studies

The importance of global environmental questions has significantly advanced the impact of climate change phenology. Whilst spatial applications continue to be a core application of phenology; in recent years the temporal dimension has also been revisited, with studies showing that temporal changes, either with a natural or an anthropogenic origin, have significantly altered phenological rhythms and seasonal development-changes attributed now to an anthropogenically induced temperature increase. This paper explores and introduces recent and newly developing analytic methods in phenology; with a view to increasing an interdisciplinary perspective and dialogue. Of particular focus is how we can and best deal with nonlinearity of phenological change in time and with multiple location studies; rigorously model the inherent multivariate time series structures in climate-phenology data; further Bayesian and non-Bayesian methods, detect multiple change-points; map seasonality calendars; model de-synchronisation of species globally; invoke old fashioned, yet rarely used circular statistical methods; adapt new transitional state modelling of phenophases with respect to climate and progress a unified paradigm for meta analytic studies in phenology. The provision of uncertainty analysis is also still much needed in climate-related phenological research. Reaching consensus on design, method of data collection and comparable analytic methods is integral to advancing the generalisability of phenological results; as is a consensus on inclusion criterion for studies selected for phenological meta-analytic studies. A coherent nomenclature is critically required, but it is currently lacking in many areas of phenology.