Vector Space Integration for Dark Matter Scattering

I present a highly efficient integration method for scattering calculations, which can reduce the evaluation time of particularly challenging analyses by a factor of \(10^7\). By projecting each item in the integrand onto vector spaces spanned by simple basis functions, a multidimensional numerical integral is replaced by a much easier exercise in matrix multiplication. This method can be generalized to any integrand that depends linearly on multiple input functions. In this paper I apply the method to dark matter (DM) direct detection with anisotropic target materials, where the DM scattering rate depends not only on the DM particle model, the astrophysical DM velocity distribution, and the properties of the target material, but now also on the \(SO(3)\) orientation of the detector. The vector space calculation factorizes the astrophysics, the DM particle model, and the SM material properties from each other, while describing detector rotations by simple matrix multiplication acting on the basis functions. Difficult analyses, which otherwise would take decades or centuries on a computing cluster, can now be performed in a few hours on a laptop.