Science Data Processing in the Cloud for MAVEN IUVS
Abstract
The Imaging UltraViolet Spectrograph (IUVS) instrument is one of several aboard the MAVEN spacecraft currently interrogating the Martian atmosphere for hints about its distant past. As the mission progresses, our total mission data volume increases. This is especially true since the MAVEN spacecraft will enjoy an extend mission life due to its new relay-satellite mode for Mars ground assets, which means that IUVS science data will continue into the foreseeable future. The data processing needs for planetary mission datasets are ever increasing, both as onboard instrumentation and computational resources expand in their capability, and as downlink bandwidth improves. As the data volume increases, the need for a high degree of parallelization becomes greater. In addition, scientific algorithms which may not have written to be easily parallelizable, need to be adapted to HPC environments so that data products can be quickly reprocessed. In particular, if an algorithm changes such that a new data product release is warranted for the entire mission dataset, fast processing of the data is highly desirable so that stakeholders can evaluate the data and make necessary adjustments or corrections. We explore cloud-based options for MAVEN IUVS data processing needs. We present a comparison of such methods to our current in-house processing infrastructure. Moving our processing into the cloud will enable us to dial-up or dial-down resources so that we can efficiently reprocess the entire mission dataset on demand. In addition, we provide a computation environment for our science team to make adjustments that can be rapidly deployed to our data processing environment.
- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFMIN11E0702E
- Keywords:
-
- 1904 Community standards;
- INFORMATICS;
- 1936 Interoperability;
- INFORMATICS;
- 1946 Metadata;
- INFORMATICS;
- 1976 Software tools and services;
- INFORMATICS