Compute trends of temperature in R with EnvStats package
Introduction Often in environmental studies we are interested in assessing the presence or absence of a long term trend. A widely applied is a parametric test for trend, which involves fitting a linear model that includes some measure of time as one of the predictor variables, and possibly allowing for serially correlated errors in the model. Instead of fitting the data to time series parametric test, Stephen Millard bundles several functions in EnvStats package that are non–parametric and agnostic in dealing with trend (Millard 2013).