NASA’s Earth Science Division hosted a vendor webinar with MDA Space on April 29. The agenda read like operations 101: discover, access, work with C-band synthetic aperture radar products. The substance underneath is a topology shift the agency has been working on for half a decade. NASA, the operator, is becoming NASA, the buyer.

The Commercial Satellite Data Acquisition (CSDA) program is the procurement layer that makes that shift official. Where NASA once funded the satellite, owned the bus, and pushed the data into its own distributed archive, it now funds an acquisition contract, owns a licence, and federates the data alongside its own missions. MDA Space — the Canadian firm that IPO’d out of MDA Ltd. in 2023 with the RADARSAT lineage and the new Chorus C-band SAR constellation — is one of the vendors on that list. From the user’s seat, you query a STAC catalogue and the data lands. From the systems-architecture seat, two hops have been added to the dependency graph: the vendor’s commercial roadmap and the procurement office’s annual budget.

←TODAY: NASA’s CSDA program hosted MDA Space at an April 29 vendor webinar — Earth observation is now a procurement line, not a capital project.
→3012: By the 3012 horizon, the practice that survives doesn’t watch satellites — it owns the real dependency graph between every satellite it rents.
Fulcrum: A national agency becomes a customer the moment it can no longer afford to be both operator and beneficiary; that transition does not reverse.

Why C-band SAR ends up on an architect’s desk

C-band SAR is the wavelength that does not care about clouds or daylight. For an AEC practice already living downstream of an Earth-observation pipeline, that has three immediate applications: persistent-scatterer InSAR over a building site before the geotechnical report comes back; dam-crest and tunnel-portal monitoring across the Alpine system; glacier-adjacent construction in the Bernese Oberland and Valais where permafrost retreat is the slow-moving constraint.

What changes when MDA Space’s Chorus constellation lands in the CSDA federation is not the physics — C-band is C-band, and ESA’s Sentinel-1 has been free since 2014. What changes is that NASA-grade preprocessing, calibration, and licensing now flow out of a commercial bus. For a Swiss office, that is a real shortening of the data pipeline. It is also a shortening that runs through a foreign commercial entity and a U.S. procurement office — two new conditional dependencies your design brief did not previously have.

Atelier: The PAZ atelier defaults to ESA’s Sentinel-1 baseline and treats commercial C-band SAR as an additive layer, not a foundation. When a vendor pivots, prices a layer out, or gets acquired, the free Copernicus baseline keeps a live project moving while you renegotiate the rest. The Practice Stack (PSS) catalogues the feed alongside the analysis — so when the supplier moves, the team knows exactly what was bought.

Hack: This Hack teaches you to pull C-band SAR scenes over a Swiss site without owning a satellite or paying a vendor. Open a notebook against Sentinel-1 — same wavelength MDA Space will sell you tomorrow, free for the asking today.

from pystac_client import Client

cat = Client.open("https://earth-search.aws.element84.com/v1")
search = cat.search(
    collections=["sentinel-1-grd"],
    bbox=[8.50, 47.36, 8.58, 47.40],   # Zurich centre
    datetime="2026-01-01/2026-05-15",
    limit=10,
)
for it in search.items():
    print(it.id, it.properties["sar:polarizations"])

Ten scenes, polarisations listed, ready for an InSAR pair pick. The query is the dependency graph in miniature: STAC API → AWS-hosted Sentinel-1 GRDs → ESA Copernicus → European taxpayer. Four hops, each one inspectable.

The warning is about topology, not technology

We did not run out of compute in my time. We ran out of intact bandwidth, intact cooling, and the institutional memory of how the federated archive worked when one of its commercial spokes went private-equity. CSDA is a healthy programme — done well, it gives NASA cadence and surface area it could never afford to launch itself. Done lazily on the buyer side, it produces a generation of analyses that quietly assume a vendor will be there next quarter.

Draw your real dependency graph this week. Not the architecture diagram — the dependency graph. For every SAR layer your office uses, write down by hand: who flies the satellite, who hosts the archive, who pays the bill, what happens if any one of those three disappears, and what your fallback is on the morning it does. The third single point of failure is the one you didn’t know you had.

Source: NASA Breaking News