Snowblind cleans cosmic rays, snowballs, and detector artifacts from JWST images, turning messy raw telescope data into pristine windows to the universe.

Every photon that reaches the James Webb Space Telescope carries precious information from across the cosmos, but Earth’s cosmic ray environment and detector imperfections can contaminate these signals with spurious hits, persistence artifacts, and the infamous ‘snowballs’ that plague infrared observations. Snowblind tackles this fundamental challenge in modern astronomy: separating real celestial signals from instrumental noise.

This Python toolkit provides four specialized algorithms that integrate seamlessly with JWST’s standard pipeline. The SnowblindStep masks cosmic ray showers and snowball artifacts with superior accuracy, while JumpPlusStep properly propagates quality flags across frame-averaged detector groups. PersistenceFlagStep identifies pixels affected by charge persistence between exposures, and OpenPixelStep detects newly failing detector elements through self-calibration techniques. Each step can run independently via command line or integrate as pipeline hooks for automated processing.

Designed for observatory pipelines and research groups processing JWST observations, Snowblind ensures that every hard-won photon from distant galaxies, exoplanet atmospheres, and stellar nurseries gets the meticulous data cleaning it deserves. Whether you’re studying the earliest galaxies or characterizing potentially habitable worlds, clean data is the foundation of groundbreaking discoveries.

⭐ Stars: 8
💻 Language: Python
🔗 Repository: mpi-astronomy/snowblind