Climate-model factor separation with Shapley values and efficient sampling
Abstract
Climate models often feature nonlinear responses to changes in model parameters or boundary conditions. Factor separation asks how much of the resulting change should be assigned to altered factors and their interactions. Lunt et al. (2010) showed that two factor attribution methods, the \emph{linear-sum} and \emph{shared-interaction} methods, coincide for n≤ 4 factors and conjectured this holds for arbitrary n. We show that the linear-sum and shared-interaction formulas are respectively the permutation and Harsanyi-dividend representations of the Shapley value. Their equality therefore follows from a classical theorem in cooperative game theory. To our knowledge, this is the first explicit identification of the Stein--Alpert and Lunt climate-model factor-separation methods with Möbius/Harsanyi coefficients and Shapley values. The result imports many established and efficient Shapley sampling methods into climate-model experimental design, and extends to vector-valued responses.
Related
- Baseflow separation combined with stratified sampling serving the accurate estimation of pollutant loads—the example of a small catchment with high anthropogenic pressures
- Effective separation of Al2O3 from coal gangue and its purification in agitation crystallization for preparation of high-quality ammonium alum
- Separation of Source, Site, and Path Effects Based on Elastic Radiative Transfer Theory. Part 1: Mapping of Scattering and Absorption Properties and Application to Metropolitan France
- Separation of Source, Site, and Path Effects Based on Elastic Radiative Transfer Theory, Part 2: Generalized Inversion of Source and Site Parameters and Application to Metropolitan France
- A simplified, non-automated, DGA-based resin protocol for the separation of Ca, Sr and REE: Validation for Ca, stable and radiogenic Sr isotopic compositions
BibTeX
@misc{jansma2026climatemodel,
title = {{Climate-model factor separation with Shapley values and efficient sampling}},
author = {Abel Jansma},
howpublished = {arXiv},
year = {2026},
month = sep,
doi = {10.48550/arxiv.2609.11948},
eprint = {2609.11948},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2609.11948},
}
RIS
TY - UNPB
TI - Climate-model factor separation with Shapley values and efficient sampling
AU - Abel Jansma
PY - 2026
DA - 2026/09/01
PB - arXiv
DO - 10.48550/arxiv.2609.11948
UR - https://arxiv.org/abs/2609.11948
AB - Climate models often feature nonlinear responses to changes in model parameters or boundary conditions. Factor separation asks how much of the resulting change should be assigned to altered factors and their interactions. Lunt et al. (2010) showed that two factor attribution methods, the \emph{linear-sum} and \emph{shared-interaction} methods, coincide for n≤ 4 factors and conjectured this holds for arbitrary n. We show that the linear-sum and shared-interaction formulas are respectively the permutation and Harsanyi-dividend representations of the Shapley value. Their equality therefore follows from a classical theorem in cooperative game theory. To our knowledge, this is the first explicit identification of the Stein--Alpert and Lunt climate-model factor-separation methods with Möbius/Harsanyi coefficients and Shapley values. The result imports many established and efficient Shapley sampling methods into climate-model experimental design, and extends to vector-valued responses.
ER -