Multidecadal Trends in Seasonal Hindcasts

Oral Presentation 

I will present a method to quantify the range of multidecadal climate trends in the Met Office’s DePreSys3 ensemble of seasonal re-forecasts (hindcasts). Dynamical models are now routinely used for prediction on seasonal to decadal timescales, but these hindcasts remain underutilised for understanding longer term climate change and variability. Each run is launched with initial conditions that reflect the observed state of the atmosphere and ocean at that time. Previous studies of hindcast trends have focused on the model ensemble-mean, neglecting the role of seasonal and subseasonal variability in generating a spread of possible trends. Over the satellite era, we find that this short-timescale variability can in fact lead to a wide range of hindcast trends in various climate metrics, such as in midlatitude jet shifts and the rate of Arctic amplification. In the forecasting world, these results suggest that caution is necessary when diagnosing seasonal forecast trend errors using only the model ensemble-mean, particularly for features that are affected by circulation variability. We also argue that hindcasts offer untapped potential for studying how short-timescale variability can affect multidecadal climate trends. This has the advantage of exploiting existing resources that are generated operationally by forecasting services, rather than requiring the bespoke generation of large climate model ensembles.