What the Land Could Power
I was looking at Google’s Project Sunroof coverage map recently — the one that shows which parts of the country have solar potential data. The map is mostly white space. Orange blobs cluster around major metros and a handful of rural counties. Everywhere else, nothing.
The white space isn’t a data problem. It’s a product decision. Project Sunroof is built to answer one question: should I put panels on my roof? That question requires very high-density LiDAR — enough to model individual roof planes, pitch, and shading from adjacent structures. The survey-grade data for that exists mostly in cities. So the coverage map looks like a metro footprint with some rural exceptions.
But there’s a different solar question — not “should I put panels on my roof” but “is this 50-acre parcel worth anything to a solar developer?” Solar siting tools exist for that question. LandGate, PVcase, Enverus PRISM — there’s a whole professional category built around it, designed for development teams running large project pipelines across dozens of sites simultaneously.
What those platforms don’t do is tell you what the trees are worth, what it actually costs to clear them net of timber harvest value, or whether a parcel’s standing forest has more long-term value than the solar project that would replace it. They assume the land question is already answered. TALON is where the land question gets answered first.
The Solar Question Listings Don’t Answer
Ground-mount and utility-scale solar development start with land. Specifically, with a land scout asking whether a particular parcel is worth pursuing before spending money on site control.
The analysis they’re doing manually — or not doing at all — is exactly what TALON is built for. Slope under ten percent. South-facing or close to it. Far enough from wetlands, protected land, and floodplains that the permitting path is clear. Close enough to a transmission line that the interconnect cost doesn’t kill the economics. Large enough to be worth the effort.
That’s a database query. It’s also a query that no general land search platform runs — not Zillow, not LandWatch, not any county GIS site a buyer would realistically use. TALON can run it in seconds across every parcel in the coverage area, and the results sit alongside timber estimates, forest quality scores, and ownership history in a single parcel bio.
The white spaces on that Sunroof map are where TALON’s solar analysis lives. Not competing with Google — operating in the territory they decided not to cover.
But What About the Trees
Asking whether a parcel is a good solar candidate is simple when the land is open. It gets complicated when it’s forested.
A lot of the land with the right slope and aspect and grid access in western NC and Vermont and rural Virginia is wooded. Some of it is scrub — recently logged second growth or early succession cover that holds minimal ecological value. Some of it is mature mixed hardwood that’s been standing for eighty years, functioning as an active carbon sink, providing watershed protection and habitat that took decades to develop.
The clearing math treats both the same. It shouldn’t.
The detection pipeline from “Counting Trees from the Sky” already distinguishes these stands, even without species identification. Even-age canopy with low height variance and smooth crown structure reads like a managed plantation. High variance, multiple canopy layers, rough crown surface — that’s structural complexity. It’s the signature of mature, unmanaged forest, and it matters.
The forest quality score that’s going into TALON isn’t a conservation statement. It’s an honest data layer. A parcel with a high solar viability score and a low forest quality score — scrub pine on a south-facing bench — is a legitimate development candidate. A parcel with high solar viability and high forest quality is a different situation. The clearing economics might still pencil. But the ecological cost is real, and the output should say so.
The four-quadrant framing is the right one: strong candidates where the economics work and the forest quality is low; controversial candidates where the economics work but the clearing cost is meaningful in ways the numbers don’t capture; marginal candidates where neither works; and preserve candidates where the forest is valuable enough that the analysis ends there.
That last category isn’t a failure of the tool. It’s the tool working.
The Clearing Math Worth Running
Here’s what makes the solar clearing analysis genuinely novel: the timber value estimates from “What the Trees Are Worth” already exist for forested parcels.
The net clearing cost for a wooded solar site is gross clearing minus timber harvest value. That number isn’t surfaced by the solar siting platforms — they’re optimizing for interconnect capacity and slope, not standing timber. A dense mature pine stand in the Carolina piedmont might return enough in stumpage to nearly fund its own clearing. A high-canopy hardwood stand might push the net cost high enough to make an adjacent open field more attractive.
The timber offset doesn’t make clearing ethically neutral. But it makes the tradeoff legible in a way it isn’t when the two analyses — solar siting and timber valuation — live in completely separate tools. TALON can close that loop — not to advocate for clearing, but to make the actual economics visible so the person making the decision has real information.
The evergreen discrimination adds a layer to this that matters. NLCD, the national land cover dataset, classifies forest type at thirty-meter resolution with a multi-year lag. It misses planted stands that postdate the last update, cedar encroachment into open fields, and small evergreen patches embedded in deciduous cover. LiDAR acquired in leaf-off season is a better discriminator — evergreen canopy maintains its density through winter, deciduous canopy nearly disappears. The first step in forest quality analysis is checking the 3DEP acquisition season for each tile. In a lot of the coverage area, the answer to “is this deciduous or evergreen” is already in the point cloud.
A Different Reason Not to Clear
While building out the forest quality model, something else emerged.
Mature mixed hardwood in the Appalachians — north-facing slopes, cove positions, 75-80 percent canopy cover — matches the habitat profile for wild ginseng almost precisely. The same structural complexity score that flags a stand as high ecological value turns out to be a reasonable proxy for ginseng habitat suitability.
Wild-simulated ginseng — seed planted in appropriate forest, left alone to develop naturally over seven to ten years — currently sells for five hundred to a thousand dollars per pound dry weight. A well-sited parcel with ideal conditions can yield twenty to fifty pounds per acre at that horizon. That’s a different conversation than the clearing economics.
The NTFP layer TALON is building isn’t a claim about what’s growing on a parcel. Wild population presence is a field question the data can’t answer. What it can answer is whether the habitat conditions are right — aspect, slope position, canopy structure, soil pH, drainage class — and what a well-managed wild-simulated operation might produce on a parcel that scores well.
Sometimes the highest-value use of a mature forested parcel isn’t solar. Sometimes it isn’t timber either. The parcel bio should say so when the numbers support it.
The Stream at the Bottom of the Ravine
I grew up on a large mountainside parcel in Vermont. There was a small stream running year-round through a ravine on the lower section. I spent a lot of time near it as a kid.
At some point I understood, in the abstract way that adolescents understand things, that the combination of that stream and the elevation drop between where it entered the ravine and where it emerged at the bottom was something you could use. You could divert some water, run it along the contour for a few hundred feet, then drop it steeply through a pipe to a small turbine. The physics are simple. Power equals head times flow times a constant. A fifty-meter drop and a couple liters per second generates roughly 700 watts continuous — enough to run a well-designed off-grid cabin indefinitely.
That stream is still running. The ravine is still there.
The head analysis for micro-hydro is a DEM problem. Find the stream channel. Trace the ravine profile. Measure the elevation difference between candidate intake and powerhouse locations. Optimize the penstock route — follow the contour to maximize head while minimizing pipe length. TALON can do all of that with the terrain data it already has. The flow estimate is harder: NHDPlus has stream network data but often nothing for the small first-order channels with the best head potential. Watershed area and regional precipitation normals give a reasonable proxy, with the honest caveat that actual low-flow measurement is required before you size a system.
What the analysis produces isn’t a construction specification. It’s a flag: this parcel has a perennial stream, meaningful head, and catchment area suggesting adequate year-round flow. The rest is fieldwork.
The reason this matters for land intelligence isn’t energy production at scale. It’s off-grid viability. A parcel with sufficient micro-hydro potential generates power continuously — not intermittently like solar, not subject to weather, not dependent on storage to bridge overnight gaps. A hundred feet of head on a small reliable stream can run a well-designed cabin indefinitely on essentially zero operating cost. That’s a fundamentally different land asset than one that’s only suitable for grid-tied solar.
The dual-generation flag — high solar viability combined with meaningful micro-hydro potential — is something no other land intelligence platform surfaces. Solar covers daytime peak demand. Hydro provides baseload continuity. Together they describe a parcel that’s genuinely self-sufficient. That’s worth putting in the parcel bio.
What the Parcel Bio Is Becoming
TALON started as a way to ask database questions about land that listing sites don’t let you ask. Acreage, terrain, ownership patterns, timber value — the information that separates a real homestead candidate from a property that looks right on paper and disappoints on arrival.
The solar and energy analysis is the next layer of that same idea. A parcel’s potential isn’t just what it is today — the timber on it, the slope, the road frontage. It’s what it could be. Whether it’s a viable solar site. Whether the forest has more value standing than cleared. Whether a ravine on the lower section could power everything on the property indefinitely.
None of this requires data that doesn’t exist. The solar radiation database is public. The stream network and watershed data are public. The terrain models that make head analysis possible are the same terrain models TALON uses for everything else. The forest structural data that distinguishes a plantation from a mature mixed stand is in the point cloud.
The land has always held the answers to these questions. The infrastructure to read them at parcel scale, across a county, and present them in one place — that’s what we’re building.
A piece of land is more than its listing. TALON is learning how much more.