A reproducible pipeline for detecting and analyzing detached haze layers in Jupiter’s atmosphere using JunoCam imagery, NAIF SPICE geometry, and ISIS3.
The Juno Haze Classification System (HCS) is an end-to-end analysis pipeline developed to identify and characterize detached atmospheric haze layers along Jupiter’s limb using data from the JunoCam instrument. The system integrates photometric brightness-gradient analysis with SPICE-based geometric constraints to distinguish primary cloud decks from secondary, optically detached haze features.
All processing is performed on a per-IMG basis, ensuring that each JunoCam image is analyzed independently with fully traceable intermediate products. This design prevents output collisions, supports reproducibility, and allows statistical aggregation across images to be performed as a distinct analytical step.
The pipeline is implemented as a sequence of modular Python stages:
- Framelet ingestion and SPICE initialization (ISIS3)
- Sub-pixel limb tracing
- Limb rectification and perpendicular sampling
- Brightness-gradient peak classification
- Latitude-binned statistical analysis of haze occurrence
Scientific conclusions are derived conservatively from aggregate statistical behavior, rather than individual limb profiles, in order to respect intrinsic geometric limitations at the planetary limb.
This repository accompanies the manuscript
“Detached Haze Classification Along the Jovian Limb Using JunoCam”
and contains the complete software and workflow required to reproduce the published results.
Two environment specifications are provided: environment.yml (portable) and env.lock.yml (exact replication of the software environment used to produce the published results).
- Fully reproducible for the JunoCam IMG(s) analyzed in the paper
- Explicit handling of SPICE geometry limitations
- Designed for extension to additional images and future studies
If you use this software or methodology in your work, please cite the accompanying manuscript and this repository.
A machine-readable citation file is provided in CITATION.cff.