Monte Carlo simulations of soft proton flares: testing the physics with XMM-Newton

TL;DR

This study uses Monte Carlo simulations to analyze soft proton flares affecting XMM-Newton, confirming their solar origin and reproducing observed proton vignetting, which helps understand and mitigate observational data loss.

Contribution

It introduces a detailed simulation framework for soft proton interactions with X-ray optics, matching real observations and clarifying the proton population's role in flare events.

Findings

Soft proton flares are influenced by Earth's magnetosphere orientation.

Simulations reproduce observed proton vignetting effects.

Proton fluxes are significantly lower than observed detection levels.

Abstract

Low energy protons (<100-300 keV) in the Van Allen belt and the outer regions can enter the field of view of X-ray focusing telescopes, interact with the Wolter-I optics, and reach the focal plane. The use of special filters protects the XMM-Newton focal plane below an altitude of 70000 km, but above this limit the effect of soft protons is still present in the form of sudden flares in the count rate of the EPIC instruments, causing the loss of large amounts of observing time. We try to characterize the input proton population and the physics interaction by simulating, using the BoGEMMS framework, the proton interaction with a simplified model of the X-ray mirror module and the focal plane, and comparing the result with a real observation. The analysis of ten orbits of observations of the EPIC/pn instrument show that the detection of flares in regions far outside the radiation belt is largely influenced by the different orientation of the Earth's magnetosphere respect with XMM-Newton's orbit, confirming the solar origin of the soft proton population. The Equator-S proton spectrum at 70000 km altitude is used for the proton population entering the optics, where a combined multiple and Firsov scattering is used as physics interaction. If the thick filter is used, the soft protons in the 30-70 keV energy range are the main contributors to the simulated spectrum below 10 keV. We are able to reproduce the proton vignetting observed in real data-sets, with a 50\% decrease from the inner to the outer region, but a maximum flux of 0.01 counts cm-2 s-1 keV-1 is obtained below 10 keV, about 5 times lower than the EPIC/MOS detection and 100 times lower than the EPIC/pn one. Given the high variability of the flare intensity, we conclude that an average spectrum, based on the analysis of a full season of soft proton events is required to compare Monte Carlo simulations with real events.