The Lexicon of Restorative Agrisolar Design
A shared design language for agrisolar photovoltaic landscapes
Restorative Agrisolar Design
Working draft, circulated for comment. Some citations are not yet verified and are marked where they appear, and the evidence base says what is established, what is inferred, and what nobody has measured.
Solar and farming want the same flat, open, well-drained ground. An array can sit on pollinator habitat or on gravel. It can slow the water leaving a field or send it somewhere new, shade a flock, end the pumping on an irrigated field, or keep a farm in business. Which of those happens is a design decision.
Agrisolar is any solar on farmland or among it, including arrays that keep a farm producing or bring farming to ground that never had it, whether or not anything grows under the panels. Agrivoltaics is the narrower case, where crops or livestock share the ground inside the fence. Growing crops directly under panels is a small part of what has actually been built, yet most agrivoltaics research is about it (Macknick et al. 2022). This lexicon is mostly about the rest, and it tries to describe them just as precisely.
Where you draw the line decides what counts. Most accounting stops at the fence, the tops of the panels, and a meter into the soil. This lexicon draws the line around the farm and the community hosting the array instead, so the lease that keeps a farm solvent, the water leaving the field, and the habitat next door all count.
Restorative agrisolar design is the idea that a solar project’s goals should include the soil, the water, the wildlife, and the neighbors, alongside the power it makes. This lexicon gives that idea words precise enough to test on a real site.
On the Oregon State University campus in Corvallis, 1.4 megawatts of solar stand on six acres of what was a non-irrigated sheep pasture, and the university’s sheep still graze it. Two peer-reviewed studies have measured what the array did. Through a dry summer the ground under the panels stayed wetter than the open pasture (Adeh et al. 2018), and in later spring trials lambs raised there grew as fast as lambs in the open (Andrew et al. 2021). Each chapter of the primer places Rabbit Hills on its own scheme, in a box like this one, so that by the last chapter the array has a full description. It illustrates the vocabulary, and it is not an audit of the project.
How to read this lexicon
The lexicon comes in two parts. Part I, the primer, is meant to be read straight through, in about half an hour. Its seven chapters each answer one question, and each is built the same way: the answer in the first paragraph, one figure, one table naming every member of the scheme, and Rabbit Hills placed on it. Part II is looked up rather than read. It holds the terminology, the full archetype catalog, the evidence base, and the sources.
| The question | The chapter that answers it | |
|---|---|---|
| 1 | What choices do designers have for shaping the array? | Design Levers: six kinds of choice |
| 2 | What forces push on the array’s design? | Design Pressures: nine forces |
| 3 | How large is the array? | The Scale Transect: five size zones, A1 to A5 |
| 4 | What happens to farming on the ground the array takes? | The Farming Axis: six positions, S0 to S5 |
| 5 | Which choices do the restorative work? | Restorative Work: six leads, one per lever |
| 6 | How does the array affect the landscape? | Restorative Functions: four functions, eight outcomes |
| 7 | What does the practice look like on real ground? | Archetypal Forms: thirteen recurring designs |
The schemes fit together in one direction. The six design levers are the choices designers have for shaping an array, and nine design pressures push on how each one gets set. Design levers do both design work, setting the form of the array, and restorative work, setting its impact. Once the levers are set, any project can be described three ways: how big it is, what happened to the farming on its ground, and which lever leads its restorative work. Those settings decide how the array affects the landscape around it, through its restorative functions. The designs that keep turning up on real ground are the archetypes.
This lexicon argues for an aim and leaves the dimensions to the site. It takes a clear position on what agrisolar is for. It does not say how high to raise the panels or how far apart to set the rows, because the same choices that restore ground in one region can wear it down in another. The same nine pressures act everywhere, but six regional factors shift their weight: sun and cloud, the grid operator, aridity, latitude, the cropping system, and irrigation. Each chapter says where its scheme bends with them. The words are meant to travel, and the right answers often do not.
How sure we are
The claims in this lexicon rest on evidence of very different strength. More grassland bird species use planted solar sites than the row-crop fields around them, and that is known because recorders listened for them through two summers (Walston et al. 2025). That a seeded ground layer keeps sediment out of the creek below is the prairie-strip result from corn and soybean fields (Schulte et al. 2017), carried over onto ground nobody has sampled under an array. The sentences making those two claims look alike. The evidence base says once which is which, so the chapters do not hedge every sentence. The largest open questions are whether several mid-sized arrays do more than one large one, which decides how much restorative work scale does, how much soil construction costs a site, and whether a restorative planting outlasts routine maintenance.
Joining in
The most useful thing anyone can send is a description of a real project, working or not. The Describe a project form on GitHub asks what the ground was before, what changed, who runs it, what was promised, what was measured, and what did not work. Without a GitHub account, send the same by email to Anthony Kendall at kendal30@msu.edu. Corrections, counterexamples, and proposed terms are welcome by email too, or as issues or pull requests on GitHub. Especially useful: an archetype that recurs in your region and is missing here, a term that lands badly with the people who have to use it, or a citation you can trace further than we have.