How Reliable Are Decadal Climate Predictions of Near-Surface Air Temperature?

Abstract

Decadal climate predictions are being increasingly used by stakeholders interested in the evolution of climate over the coming decade. However, investigating the added-value of those initialised decadal predictions over other sources of information typically used by stakeholders generally relies on forecast accuracy, while probabilistic aspects, although crucial to users, are often overlooked. In this study, the quality of the near-surface air temperature from initialised predictions has been assessed in terms of reliability, an essential characteristic of climate simulation ensembles, and compared to the reliability of non-initialised simulations performed with the same model ensembles. Here, reliability is defined as the capability to obtain a true estimate of the forecast uncertainty from the ensemble spread. We show the limited added-value of initialisation in terms of reliability, the initialised predictions being significantly more reliable than their non-initialised counterparts only for specific regions and the first forecast year. By analysing reliability for different forecast system ensembles, we further highlight the fact that the combination of models seems to play a more important role than the ensemble size of each individual forecast system. This is due to sampling different model errors related to model physics, numerics and initialisation approaches involved in the multi-model, allowing for a certain level of error compensation. Finally, this study demonstrates that all forecast system ensembles are affected by systematic biases and dispersion errors that affect the reliability. This set of errors makes bias correction and calibration necessary to obtain reliable estimates of forecast probabilities that can be useful to stakeholders.

Representing Model Uncertainty in Multiannual Predictions

“…Results for the MME indicate a reliable ensemble over most of the globe and especially over large areas of the tropical Pacific Ocean up to the third winter. These findings are in agreement with previous studies that have demonstrated the improved reliability of multimodel compared to SMEs (Doblas‐Reyes et al., 2009; Palmer et al., 2004; Stockdale et al., 2009; Verfaillie et al., 2020; Weisheimer et al., 2009). To compare the level of reliability of all three ensembles quantitatively, we analyzed a number of grid cells which are significantly unreliable over the tropical band (20°N–20°S) (Table ).…”

“…In particular, it was demonstrated for seasonal forecasts that a fixed‐size ensemble constructed from different models is more reliable than an ensemble of the same size from a single model (Palmer et al., 2004; Stockdale et al., 2009; Weisheimer et al., 2009). Similar results comparing MMEs versus single‐model ensembles (SMEs) were recently found for climate predictions and projections on decadal time scales (Verfaillie et al., 2020).…”

NorCPM1 and its contribution to CMIP6 DCPP

“…2.2.1) (Meehl et al, 2021). A better understanding of the three aspects -initialization, model dynamics and forcing responses -is fundamental for better exploiting the climate predictive potential and improving estimates of climate predictability (Keenlyside and Ba, 2010;Cassou et al, 2018;Verfaillie et al, 2021). The existing climate prediction systems undersample effects of model and initialization uncertainty and are not necessarily well suited to address questions related to changes in the observing system.…”

Climate Services Toolbox (CSTools) v4.0: from climate forecasts to climate forecast information

“…6). This function can be used to quickly compare modes of variability, skill metrics, differences between methods, or forecast systems as a function of the lead times or seasons (Torralba, 2019;Verfaillie et al, 2021;Lledó and Doblas-Reyes, 2020).…”