(Monty Python and the) HoliGRALE: A Hybrid GRALE Lens Inversion Methodology Diagnostically Evaluated with Synthetic and Real Data

We present HoliGRALE (Hybrid, only light included, GRAvitational LEnsing), a hybrid lens modeling extension of the freeform method GRALE. HoliGRALE incorporates physically motivated parametric components within the freeform genetic algorithm GRALE. Using three synthetic clusters from IllustrisTNG, each of varying complexity, we evaluate its performance according to its accuracy in reconstructing the density distribution, image positions, magnifications, and time delays. We find that HoliGRALE generally outperforms GRALE on all fronts, and shows greater accuracy in regions containing a high image number density. We estimate the error budget in projected density to be <5% for simple, few component clusters, and about5-10% for complex, multi-component clusters in the regions constrained by observed images. We then use HoliGRALE to model three lenses: RX J2129.7 + 0005, MACS J0138.0 - 2155, and Abell S1063. In each case, HoliGRALE is able to accurately reconstruct observed magnifications and time delays. In its predictions, HoliGRALE is found to be more precise than GRALE. We particularly highlight the predicted time delay for the next arriving image of SN Requiem to be 4346.61+490.27-818.36 days (about April 2026 -- October 2029), consistent with previous work, albeit with less precision. In Abell S1063, we report the reconstruction of a main dark matter halo core radius of 89.9 plus or minus33.2 kpc, in excellent agreement with past work and useful in testing theories of dark matter. Our results emphasize the expected error budget of HoliGRALE, which should be mentioned and scrutinized in future applications of HoliGRALE for other lens systems.