The updated Multi-Model Large Ensemble Archive and the Climate Variability Diagnostics Package: new tools for the study of climate variability and change

Observations can be considered as one realisation of the climate system that we live in. To provide a fair comparison of climate models with observations, one must use multiple realisations or ensemble members from a single model and assess where the observations sit within the ensemble spread. Single model initial-condition large ensembles (LEs) are valuable tools for such an evaluation. Here, we present the new Multi-Model Large Ensemble Archive (MMLEAv2) which has been extended to include 18 models and 16 two-dimensional variables. Data in this archive have been remapped to a common 2.5 × 2.5° grid for ease of inter-model comparison. We additionally introduce the newly updated Climate Variability Diagnostics Package version 6 (CVDPv6), which is designed specifically for use with LEs. The CVDPv6 computes and displays the major modes of climate variability as well as long-term trends and climatologies in models and observations based on a variety of fields. This tool creates plots of both individual ensemble members, and the ensemble mean of each LE, including observational rank plots, pattern correlations, and root mean square difference metrics displayed in both graphical and statistical output that is saved to a data repository. By applying the CVDPv6 to the MMLEAv2, we highlight its use for model evaluation against observations and for model intercomparisons. We demonstrate that for highly variable metrics a model might evaluate poorly or favourably compared to the single realisation the observations represent, depending on the chosen ensemble member. This behaviour emphasises that LEs provide a much fairer model evaluation than a single ensemble member, ensemble mean, or multi-model mean. By leveraging the combination of the CVDPv6 and MMLEAv2 presented in this paper we can determine which climate variables need a LE for fair assessment against observations.