Abstract

We describe a custom outlier rejection algorithm for JWST/NIRSpec integral field spectroscopy. This method uses a layered sigma clipping approach that adapts clipping thresholds based upon the spatial profile of the science target. We find that this algorithm produces a robust outlier rejection while simultaneously preserving the signal of the science target. Originally developed as a response to unsatisfactory initial performance of the jwst pipeline outlier detection step, this method works either as a standalone solution, or as a supplement to the current pipeline software. Comparing leftover (i.e., not flagged) artifacts with the current pipeline's outlier detection step, we find that our method results in one fifth as many residual artifacts as the jwst pipeline. However, we find a combination of both methods removes nearly all artifacts-an approach that takes advantage of both our algorithm's robust outlier rejection and the pipeline's use of individual dithers. This combined approach is what the TEMPLATES Early Release Science team has converged upon for our NIRSpec observations. Finally, we publicly release the code and Jupyter notebooks for the custom outlier rejection algorithm.