[Discussion] INT-20-2b Feedback: Support for RGEs

[ckoerber]

This issue summarizes the feedback given at the INT-20-2b program: Beyond-the-Standard-Model Physics with Nucleons and Nuclei.

Slides (strops starting at no. 26), Video (demo starting at ~ 35min mark)

RGEs & Scale dependence of quark matter operators

Vincenzo Cirigliano emphasized the Renormalization Group evolution of interactions at the ~ 42:20 mark. Paraphrasing his comment:

Depending on the scale you work on, the quark-operator relations may change. For example, when going from the weak scale to the QCD scale. This dependence, different from current low-energy structures, is more dynamical/continuous and is described by the RG evolution of DM interactions (see also works by Joachim Brod et al.). If it were possible to implement this dynamic, high-energy scale dependence as well, this would allow running global DM fits also including direct detection experiments (maybe already possible) and collider constraints.

SM & Beyond libraries

In this context, @evanberkowitz pointed out possible connections to:

• https://github.com/flav-io/flavio
• https://github.com/wilson-eft/wilson
• https://github.com/smelli/smelli

It may be possible to connect APIs, utilize these tools to extend accessible scales and allow a more extensive set of observables to be connected.

Next steps

Literature research: What is needed, what is feasible and what is done already to establish this connection?

[ckoerber]

After some further literature research:

• Brod et. al. already provide a Python (and Mathematica module) which does the RG running down to the nuclear scale and allow to connect to the Mathematica notebook developed by Wick and collaborators. Here is the inspire query to related works by Bishara and Brod et. al.
• Reverse search of this package brought me to potentially related works:
  • A DAMA/Libra-phase2 analysis in terms of WIMP-quark and WIMP-gluon effective interactions up to dimension seven
  • Limits on Dark Matter Effective Field Theory Parameters with CRESST-II
  • Assessing Near-Future Direct Dark Matter Searches with Benchmark-Free Forecasting
  • Identifying WIMP dark matter from particle and astroparticle data
  • Improved Limits on Scattering of Weakly Interacting Massive Particles from Reanalysis of 2013 LUX Data

In this context particularly relevant seem:

• A combined analysis of PandaX, LUX, and XENON1T experiments within the framework of dark matter effective theory
• And a workshop Statistical challenges in the search for dark matter which also points to GAMBIT: The Global and Modular Beyond-the-Standard-Model Inference Tool. Their conclusion is

The detection of new physics amounts to a challenge of model selection, which comes with a number of pitfalls: the construction of a good test statistic, likelihood function or detection criterion; the choice of physical model parameters and their priors; and the efficient exploration of that parameter space. [...] Combining experiments and approaches poses additional challenges, as nuisance parameters may or may not overlap, and systematic uncertainties may have varying effects. Approaches such as hierarchical modeling and global fitting help tackle these daunting issues with finite effort.

They provide more details in section IV.B.2.

Care needs to be taken when combining data from different experiments. In particular, consistent theoretical calculations and assumptions must be applied toall predictions (and even experimental likelihoods) in a global fit, along with the same assumptions about Standard Model parameters, nuclear physics and astrophysical aspects like the density and velocity distributions of DM [293–295]. Similarly, different theoretical calculations and experimental analyses must be invoked for different theories, alternative theoretical calculations should be considered for the same model in order to estimate theory errors accurately, sophisticated statistical sampling schemes must be employed for analysing high-dimensional parameter spaces [296–298] and cover-age properties need to be checked carefully [299–301]. Ideally, the global fitting framework itself should automatically ensure that all of these requirements are satisfied. This is something that has only recently become possible [302] (this points to GAMBIT); future work is focussed on extending this rigorous consistency-checking and automation as far as the Lagrangian level.

But also

Once a new DM model is proposed, an initial choice of priors on its parameters must be made, based on some theoretical constraints or preferences of the researcher. If the likelihoods are strong, the final fitted result will not be dependent on the chosen prior. However, because only the Planck [303] relic density measurement provides a clear “signal” amongst the different DM experiments,the likelihoods for DM searches are very often too weak to dominate over the impact of the prior on the posterior.

They further point to

• LikeDM: likelihood calculator of dark matter detection
• A generic method to constrain the dark matter model parameters from Fermi observations of dwarf spheroids
• Possible dark matter annihilation signal in the AMS-02 antiproton data
• AMS-02 Positron Excess and Indirect Detection of Three-body Decaying Dark Matter

This paper also contains a toy example for combining experiments.

[ckoerber]

It seems like a significant part of the intended work was published in Thermal WIMPs and the Scale of New Physics:Global Fits of Dirac Dark Matter Effective Field Theories.