# JPEG2000 Image Compression on Solar EUV Images

## TL;DR

This paper evaluates the effectiveness of JPEG2000 lossy compression on solar EUV images from SDO/AIA, analyzing image quality and scientific usability for space missions and large data archives.

## Contribution

It provides an assessment of JPEG2000 compression effects on solar images, including quality metrics and validation for scientific analysis.

## Key findings

- JPEG2000 compression maintains acceptable image quality at certain bit rates.
- Lossy compression can be scientifically valid for analyzing solar phenomena.
- Comparison of MSSIM and PSNR metrics for solar EUV images.

## Abstract

For future solar missions as well as ground-based telescopes, efficient ways to return and process data have become increasingly important. Solar Orbiter, e.g., which is the next ESA/NASA mission to explore the Sun and the heliosphere, is a deep-space mission, which implies a limited telemetry rate that makes efficient onboard data compression a necessity to achieve the mission science goals. Missions like the Solar Dynamics Observatory (SDO) and future ground-based telescopes such as the Daniel K. Inouye Solar Telescope, on the other hand, face the challenge of making petabyte-sized solar data archives accessible to the solar community. New image compression standards address these challenges by implementing efficient and flexible compression algorithms that can be tailored to user requirements. We analyse solar images from the Atmospheric Imaging Assembly (AIA) instrument onboard SDO to study the effect of lossy JPEG2000 (from the Joint Photographic Experts Group 2000) image compression at different bit rates. To assess the quality of compressed images, we use the mean structural similarity (MSSIM) index as well as the widely used peak signal-to-noise ratio (PSNR) as metrics and compare the two in the context of solar EUV images. In addition, we perform tests to validate the scientific use of the lossily compressed images by analysing examples of an on-disk and off-limb coronal-loop oscillation time-series observed by AIA/SDO.