The NSF National Geophysical Facility’s seismological repository is transitioning into the cloud, and data submitters and consumers should be aware of the following operational changes:

1. 1. New data submissions take precedence and replace previously submitted data.
2. 2. Quality codes in miniSEED records will no longer be modified.

When

These changes will apply to all miniSEED data submitted to the facility after the cloud-based archiving system becomes operational sometime in the first half of 2026. The unmodified quality codes are already visible for data distributed from real-time collection buffers.

1. Most recently submitted data takes precedence

Archiving miniSEED data involves processes that prepare it for the long-term repository and general availability. These include sorting and removing overlapping data to form the most continuous time series possible organized in a way ready for processing and distribution.

What changed

The new miniSEED archiving system prioritizes newly submitted data based on identifiers and time coverage, replacing any existing overlapping data in the repository. Previously, the miniSEED quality codes were included in the determination of priority.

Removal of miniSEED data is not automated and remains a manual process. Data contributors who wish to request data removal should contact help@earthscope.org.

Furthermore, data that are not streaming in (near) real-time should be submitted using the facility’s batch dropoff mechanism, even if replacing data originally submitted as a stream.

2. Quality codes will no longer be modified

The legacy miniSEED archiving process sets the data quality code for each record to “M”’ to distinguish it from submitted data. This modification indicates that the records had potentially been modified by the facility and helps determine which records to prioritize during updates. Unfortunately, this approach required the facility to modify every record submitted and did not allow contributor-designated quality codes to reach users.

What changed

Quality codes are no longer modified by the new archiving process and are distributed without change to users. If data are submitted with multiple qualities, users may receive data for a single time series with multiple qualities.

The “M” quality data already in the repository will retain that code.