Climate Model Code Genealogy and Its Relation to Climate Feedbacks and Sensitivity

Abstract

Contemporary general circulation models (GCMs) and Earth system models (ESMs) are developed by a large number of modeling groups globally. They use a wide range of representations of physical processes, allowing for structural (code) uncertainty to be partially quantified with multi‐model ensembles (MMEs). Many models in the MMEs of the Coupled Model Intercomparison Project (CMIP) have a common development history due to sharing of code and schemes. This makes their projections statistically dependent and introduces biases in MME statistics. Previous research has focused on model output and code dependence, and model code genealogy of CMIP models has not been fully analyzed. We present a full reconstruction of CMIP3, CMIP5, and CMIP6 code genealogy of 167 atmospheric models, GCMs, and ESMs (of which 114 participated in CMIP) based on the available literature, with a focus on the atmospheric component and atmospheric physics. We identify 12 main model families. We propose family and ancestry weighting methods designed to reduce the effect of model structural dependence in MMEs. We analyze weighted effective climate sensitivity (ECS), climate feedbacks, forcing, and global mean near‐surface air temperature, and how they differ by model family. Models in the same family often have similar climate properties. We show that weighting can partially reconcile differences in ECS and cloud feedbacks between CMIP5 and CMIP6. The results can help in understanding structural dependence between CMIP models, and the proposed ancestry and family weighting methods can be used in MME assessments to ameliorate model structural sampling biases.

KNMI'23 Climate Scenarios for the Netherlands: Storyline Scenarios of Regional Climate Change

“…We therefore decided to take into account all CMIP6 models that had the required data available, and not to exclude specific models on the basis of (subjective) criteria related to the model's ability to reproduce the historical climate. Similarly, we decided not to remove models that share certain aspects of the model code (e.g., code for particular parameterizations of physical processes, or the code for the specific model sub‐components) and are therefore not completely independent (e.g., EC‐Earth3 p1 and CNRM in the wet‐trending group, ECMWF family; UKESM1‐0 and ACCESS‐ESM1‐5, HadAM3 family; see Kuma et al., 2023).…”

Modelling temperature‐precipitation pressures on African timber heritage

“…Models included within this study were required to have: (i) daily precipitation and 2 m surface temperature (Brimblecombe & Richards, 2023) and (ii) outputs for the years 1984–2013 (past); 2035–2064 (mid‐century); 2070–2099 (end‐century). Where multiple models were available from the same modelling group, only one model was retained for analysis (typically the model with higher spatial resolution) to reduce the risk of model structural biases introduced by model interdependencies (Knutti et al, 2013; Kuma et al, 2023). Thus, we included 13 models in our analysis (Table 1).…”

Assessment of Precipitation and Hydrological Droughts in South America through Statistically Downscaled CMIP6 Projections

“…Given their equal origin, there may be a higher likelihood of shared biases or underlying assumptions, leading to correlated errors and limited diversity within the ensemble. In this context, the "institutional democracy" approach addresses these uncertainties by selecting one GCM from each modeling institute [101,102]. While this is an effective way to account for model dependence, it is worth noting that as institutes progressively copy or collaboratively develop models or components, there is no guarantee that such an approach will continue to be efficient in future studies [103].…”