Free Radar Makes the Model the Moat

Radar imagery of nearly every piece of dry land on Earth is now free, and that is exactly why it is no longer where the money is. What energy and mining companies will pay for is the thing the imagery has to be pinned to: the utility's network model, the miner's survey-grade terrain model.

Start with the giveaway. NASA's Jet Propulsion Laboratory said the U.S. side of the NISAR mission began releasing calibrated products continuously on July 20, covering all L-band measurements collected since June 17. The satellite, launched July 30, 2025, monitors nearly all land- and ice-covered surfaces on a 12-day cycle, according to JPL. Its longer-wave L-band can pass through tree canopies and image the ground beneath, which matters for anyone watching a pipeline right-of-way through forest. The output is vast: "on the order of dozens of terabytes a day," JPL said. By the end of the year the team expects to have released everything acquired earlier in science operations.

ISRO, the mission's Indian partner, lists a "Free and open data policy" in the payload description. Among NISAR's stated aims is characterizing subsidence and uplift tied to aquifers and hydrocarbon reservoirs. In other words, it is aimed squarely at energy assets, and anyone can download the results.

What a free input does to a market

Project Geospatial, in an analysis by Adam Simmons, puts the global SAR market at roughly $4.05 billion in 2025, rising to over $10.44 billion by 2034. The same piece is blunt about where that value will pool: "raw Earth observation data is becoming a commoditized input, while the real financial moats are being built at the application and intelligence layers." It says free government missions, NISAR included, leave commercial providers at "severe risk of commoditization" unless their resolution, revisit rates or analytics clearly beat the public data. It even floats a "NISAR effect" that could push commercial operators out of agriculture and environmental markets to chase defense and intelligence work.

The survivors are selling something other than pictures. Project Geospatial notes that InSAR, the technique for detecting millimeter-level ground subsidence, once demanded months of manual processing by geospatial engineers. LiveEO and Capella Space now offer it as automated, cloud-based platforms. LiveEO's SurfaceScout charges pipeline operators recurring software fees for continuous, AI-derived encroachment alerts. Once that stream feeds a customer's maintenance ERP, Project Geospatial says, switching costs are "immensely high." The lock-in comes from the integration, not from the photons.

The utility's model is the asset

Now look at who owns the layer beneath the imagery. Esri's ArcGIS Blog, in a post dated May 14, 2026, says ArcGIS Pro 3.7 and ArcGIS Enterprise 12.1 are the 2026 network management release for its Utility Network. Esri says the aim is a stable, long-term-support platform with a predictable upgrade schedule. That sounds dull, and it is the point. A utility does not rip out its network model the way it cancels an imagery subscription.

The release is also built for plugging into other systems. Trace results can now carry flow direction and propagated values in output JSON, a feature Esri calls "designed to support integrating the network model with external systems." Utility Network version 3 is being deprecated, and customers are told to move to version 5 or later.

Dominion Energy Virginia shows what the migration costs. UDC says Dominion began moving its GIS to the Utility Network in 2024 on a 3-year strategy, spanning 5 databases: electric distribution, substation and transmission, plus privatized gas and electric. UDC's agenda for the Infrastructure Management & GIS conference says Dominion is "nearing completion of its ArcGIS Utility Network migration" and will launch an upgraded digital lifecycle tying together design, as-builting and GIS. A free radar feed can tell a utility where the ground is moving. Only a clean network model can tell it which customers sit downstream.

The miner's terrain model

The same logic applies in a Malawi rutile deposit. Fortuna Metals (ASX: FUN) said in an announcement dated 21 October 2026, relayed by Discovery Alert, a partner-produced feature, that a LiDAR drone survey covering 180km² at its Mkanda project is due for completion in October 2026. It will deliver a bare-earth terrain model with X, Y and Z coordinates at 1m x 1m spacing.

Discovery Alert says a high-resolution terrain model is required to model resources to Indicated category, so it gates the upgrade of the maiden inferred estimate announced 15 July 2026. CEO Tom Langley put it plainly, per Discovery Alert: "The Lidar is critical for the next resource upgrade to Indicated." Resource models place drill results in three dimensions, Discovery Alert explains, so an accurate surface keeps the mineralisation in the right place relative to the ground. Fortuna has tightened its drill spacing to 200m x 200m from 400m x 400m in 2025, which makes a precise surface worth more, not less. Free satellite radar does not substitute for that model, because the resource estimate rests on it.

The counter-argument

The main objection is that commoditization has limits, and some sensing businesses will thrive. Project Geospatial's competitor table lists ICEYE with more than 70 X-band satellites and 16cm resolution on its Gen4 craft, far finer than anything a free mission offers. It also says defense and intelligence buyers account for over 40% of the SAR market. Sharp commercial sensing survives, it argues, by selling to people for whom resolution is the product.

Conceded. But that supports the thesis: the commercial survivors are retreating up-market toward defense, away from the utility and resource work NISAR targets. And the same analysis warns that scaling pipeline and utility monitoring "requires overcoming conservative procurement cycles within heavily regulated utility sectors." Slow procurement makes the incumbent model stickier, not weaker.

One caution on the free feed itself. NASA Earthdata says a migration will interrupt data ingestion at ASF DAAC, the Fairbanks archive that distributes NISAR files, among other centers, on Oct. 28-30, 2026. Free is not the same as frictionless.

What follows

If this read is right, the next round of geospatial value goes to whoever holds the authoritative model: the utility that finishes its migration, the explorer that owns the survey, the vendor whose platform sits inside the maintenance system. Everyone else rents the same free radar. Fortuna's drones are due to finish in weeks. The imagery overhead costs nothing; the model under it is what Fortuna is flying 180km² to build.