lenscape: Fast, High-resolution $Δχ^2$ Maps of Microlensing Events

Abstract

Planetary microlensing analyses rely on $Δχ^2$ maps across the binary-lens separation $s$ and mass ratio $q$ to locate degenerate solutions and bound parameter uncertainties. Because every grid point requires a full light-curve optimization with finite-source magnifications, constructing these maps is often the primary computational bottleneck in event modeling. Here I present lenscape, an open-source Python package that accelerates $Δχ^2$ mapping and suppresses local-minimum artifacts through multi-stage neighbor propagation and adaptive contour refinement. Applied to the planetary event MOA-2022-BLG-091, lenscape reproduces the published $50\times50$ map in 14 minutes and generates a $400\times400$ map in under two hours on a desktop computer.