Dynamic Wisp Removal in JWST NIRCam Images

Abstract

The James Webb Space Telescope (JWST) near-infrared camera has been found to exhibit serious wisp-like structures in four of its eight short-wavelength detectors. The exact structure and strength of these wisps is highly variable with the position and orientation of JWST, so the use of static templates is non-optimal. Here we investigate a dynamic strategy to mitigate these wisps using long-wavelength reference images. Based on a suite of experiments where we embed a worst-case scenario median-stacked wisp into wisp-free images, we define suitable parameters for our wisp removal strategy. Using this setup we re-process wisp-affected public Prime Extragalactic Areas for Reionization and Lensing Science data in the North Ecliptic Pole Time Domain Field, resulting in significant visual improvement in our detector frames and reduced noise in the final stacked images.