In Situ Characterization of Beta Scattering at TRINAT

Recently, the TRINAT collaboration reported a 0.32\% measurement of the $\beta$ asymmetry parameter, $A_\beta$, in the isobaric analgue decay of polarized $^{37}$K [PRL {\bf 120} (2018)].  The statistical and systematic uncertainty components are equal at 0.23\%, with the largest systematic sources arising from backgrounds (0.14\%) and $\beta$ scattering (0.12\%).  We are able to test our GEANT4 modelling of $\beta$ scattering using a small subset of the data which we veto in the main analysis:  events in our $\beta$ telescopes which register two separate pixels in the double-sided Si-strip detector along with a signal in the plastic scinitillator are predominantly due to $\beta$s which backscatter out of the scintillator. 

We will show how well GEANT4 reproduces the scintillator backscatters and discuss its impact on the $A_\beta$ measurement.  $\beta$ scattering currently limits our precision to 0.12\% and is expected to be our dominant source in the next experiment with $^{37}$K.  We will discuss our plans to adapt our geometry to reduce scattering effects and how to better benchmark our simulations to improve this source of uncertainty, allowing us to reach better than 0.1\% precision.