There is still a federal regulation that defines desert land by what it cannot do.
Under the rules implementing the Desert Land Act of 1877, public land qualifies as "desert" if it "will not, without artificial irrigation, produce any reasonably remunerative agricultural crop by the usual means or methods of cultivation." That is half the legal test — the other half is that the land be irrigable. Not what lives there. What it fails to grow for us. The Act is still on the books — 320 acres, four years to bring water onto it and cultivate an eighth of it — and the verb the statute uses for that is reclaim, as if the desert were something taken by mistake and owed back.
A desert is not land that failed to become a farm. It is a system that works because it is dry, and most of the work happens in the hours and weeks nobody is standing there to watch.
what a desert actually is
Drylands are the family. A desert is the dry end of it.
The working definition is a ratio, not a mood. The aridity index divides mean precipitation by potential evapotranspiration — what the sky delivers, over what the atmosphere would take if it could. Land scoring 0.65 or lower is dryland: the air wants at least half again more water than it gets. Inside that family sit four classes — dry subhumid, semi-arid, arid, and hyper-arid. Desert is the arid and hyper-arid end, where the index falls below 0.2.
That family covers 40.6% of the world's land outside Antarctica, and roughly 2.3 billion people live in it — about one in three humans, double the dryland population of 1990. This post reads the North American deserts; the same test holds in the Sahel or the Thar. Not the grocery-access sense of the word. Drylands are not the margin of the map. They are a third of it, with a third of us on them.
Strip away the postcard and a functioning desert is four things working at once.
- Perennial cover playing a very long game. Creosote, bursage, mesquite, saltbush, ocotillo, cholla, saguaro, Joshua tree. These are not sparse because the system is failing; they are spaced because water is the limiting input and roots claim territory accordingly. A creosote ring in the Mojave known as King Clone has been estimated at roughly 11,700 years old. The plant you are stepping around may predate agriculture.
- A living soil surface. In dry country the top inch is often a living crust — the dark, brittle skin on an undisturbed bajada that knits the surface so a monsoon soaks in instead of sheeting off. The crust that holds the desert together owns that layer.
- A wash network. Arroyos and washes are dry most of the year and are not therefore idle. They carry the monsoon, concentrate water into the strips of denser vegetation that carry most of the desert's wildlife, slow flood peaks before they reach anything expensive downstream, and recharge what is underneath.
- Timing. Desert species are keyed to a pulse rather than a season. A saguaro grows one to one and a half inches in its first eight years, under the shade of a nurse tree — usually a palo verde, ironwood, or mesquite. Arms appear at 50 to 70 years, up to 100 in drier country. Adults run 150 to 200 years. Its stem growth happens in July, August, and September, on the monsoon. Its flowers open after dark for bats and moths.
A desert is not a place where life failed to arrive. It is a place where life solved for water and stayed. Every one of those four parts is invisible at highway speed, and three of them are invisible at any speed unless you are looking for them.
empty land, a failed forest, a solar pad, or a desert
Four different things can be said about the same arid acre. Only one of them describes what is there.
| what is actually there | who manages it | how it comes back | what it produces | |
|---|---|---|---|---|
| empty land | A wash, a bajada, and specialist cover that a BLM RMP already maps — called leftover because nobody put a use on it. | A rancher, a Tribe, an agency, or weather. The vacancy is in the appraisal, not on the ground. | The label does not grow back. The dryland under it can, slowly. | Dust control, shade, and a place that stays itself until the next lease. |
| a failed forest | a desert, measured against a biome it was never going to be | a scoring system, not a manager | the comparison never resolves, because nothing was lost | a deficit on paper |
| a solar pad | array, transmission tie-in, graded and fenced ground | an operator, on a lease and a decommissioning clock | as a pad; the dryland underneath is a separate question | megawatts, which are real |
| a living dryland | perennial cover, soil crust, wash network, specialist species | ecological management — or, too often, nobody with a mandate | years to centuries, depending on which part you broke | habitat, infiltration, dust control, flood attenuation, shade, a place that stays itself |
Two of those rows have their own posts, because both deserve more than a table cell. Planting a forest on a desert is the subject of greening the desert is not restoring a desert. Utility-scale solar — which belongs somewhere, and often belongs on already-disturbed ground — is the subject of a solar pad is not a desert. Neither one is a villain. The failure mode in both is the same: calling the replacement the thing.
the wasteland was a legal category, not an observation
This matters because the "empty" reading is not a vibe. It was written down, funded, and administered.
Nineteenth-century American land policy assumed the highest use of arid land was irrigated agriculture, and built statutes to get there: the Desert Land Act in 1877, the Reclamation Act in 1902, and a whole vocabulary — reclaim, waste, unimproved — in which a working desert was a problem awaiting correction. This was not stupidity. It was a coherent policy for a country settling land that already had people and calendars on it, and the families who filed those entries were doing exactly what the law asked of them. But policy vocabulary outlives policy. Land officially described as producing nothing gets carried at what it produces on paper, which is nothing, and land carried at nothing is the first thing anyone reaches for.
Not every calendar in that country counted the same way. The Tohono O'odham — whose name translates as Desert People — open their year with Ha:ṣañ Bak Maṣad, the saguaro ripening moon, harvesting bahidaj in late June and carrying it into the rain ceremony that meets the monsoon. A year that begins with the desert producing is a different piece of accounting than a statute defining the desert by what it will not produce. That relationship is theirs to describe, and they do. The point here is narrower: "unproductive" was always a choice about what to count.
four deserts, not one desert
North America has four, and they behave differently because their rain arrives at different times of year.
- Great Basin — cold desert, precipitation mostly as winter snow, sagebrush and shadscale. Its shrub-steppe sits on the grassland side of the ledger; the sagebrush clock is sagebrush takes decades, and the empty-land cousin is what grasslands actually are.
- Mojave — winter rain, the Joshua tree's range, the transition between cold and hot desert.
- Sonoran — two rainy seasons, winter frontal storms plus the summer monsoon, which is why it carries saguaro, palo verde, and ironwood and reads as green to visitors expecting sand.
- Chihuahuan — the largest hot desert in North America, mostly above 1,000 m, mostly summer rain, creosote and lechuguilla and yucca.
Same spine, four dialects: the type determines what the work has to look like. The full comparison, plus the Baja fog-influenced fifth case and the live place agents behind each, is post five: sonoran, mojave, chihuahuan, and great basin are not the same desert.
a degraded dryland and a living desert are not the same problem
Worth separating, because the words collide.
When a dryland stops working, that is desertification — productive land losing water function and cover — and the desert isn't destiny owns that repair.
This post is the opposite failure. A healthy desert read as degraded land because it is dry. A Sonoran bajada is not a broken grassland. A creosote flat is not a field that gave up. Confusing the two is how a repair budget turns into a conversion budget, and how "restoration" ends up meaning irrigation.
what a living dryland pays for
None of it is invoiced. All of it gets paid for when it stops.
- Erosion and dust control. Painter and colleagues found that dust blown off disturbed Colorado Plateau desert soils darkened snow in Colorado's San Juan Mountains enough to shorten snow cover duration by 18 to 35 days. The soil surface of a desert is a snowmelt-timing asset for a watershed hundreds of miles away, and nobody bills for it.
- Water. Washes infiltrate and recharge; intact upland cover slows what reaches them. The alternative to a functioning wash is a concrete channel with a maintenance budget.
- Risk resilience. Flood attenuation on the monsoon, and a longer fire interval than the same ground carries once invasive annual grasses arrive.
- Habitat. Desert tortoise, bats that pollinate columnar cactus, migratory birds routed through washes and springs, pollinators keyed to bloom timing that exists nowhere else.
- Shade, surface temperature, and air. Vegetated desert ground runs cooler than bare or paved desert ground, which is a public-health variable in every desert city in the Southwest.
- Research, medicine, and legacy. Organisms that solved extreme water and heat stress are a genetic library, and the desert is a cultural one.
In the stocks and flows accounting we use, Desert is one of 15 ecosystem stocks and those are named flows off it. That framing exists for one reason: to make a living dryland legible to people who allocate capital, so the acre can be funded before it is converted.
the desert is the asset
The desert exists whether or not anyone buys a certificate. It was there before the accounting, it does not need the accounting to be real, and it will outlast whatever instrument we wrap around it. The only job the instrument has is to get someone paying for the dryland's condition while the dryland is still in condition.
Plenty of people are already paying for the other outcome. Water utilities pay for dust and sediment. Cities pay in heat and emergency response. Counties pay for flood infrastructure where a wash used to do the work. Developers pay for habitat mitigation after the siting decision instead of before it. Land trusts pay to buy back what was cheap because it looked empty. Every one of those is a bill for a dryland treated as leftover.
Mechanically, funding the other side is simple enough to state in one sentence: a certificate funds the condition of one named place directly, a coin funds protection across the protocol indirectly, and both route proceeds to an account that represents a place or a purpose. That is the entire technical content of this post. The mechanism is ensurance — funding a living dryland's condition on a named place.
Our own stage, stated plainly: deserts.ensurance is live, alongside named place agents for the Sonoran, Mojave, Chihuahuan, and Great Basin shrub-steppe. Volumes are small. The mechanism works and the market for it is early, which is a different sentence from "this is solved."
And the price is a bridge, never the worth. A desert is not worth a solar lease, a mitigation credit, or a dust-control line item. Those numbers exist so that capital can see the dryland at all — which is the whole problem a 150-year-old statute created by defining the place as the absence of a crop.
frequently asked questions
what is a desert?
A desert is an ecosystem at the dry end of the dryland spectrum — technically, land with an aridity index below 0.2, meaning potential evaporation vastly exceeds precipitation. Functionally it is a structured living system: long-lived perennial plants spaced by water availability, a biological soil crust holding the surface, a network of ephemeral washes that move and store the year's rain, and species whose life cycles are keyed to rainfall pulses rather than to seasons. A desert is dry by definition and alive by observation.
what are drylands?
Drylands are all land with an aridity index of 0.65 or lower — precipitation divided by potential evapotranspiration. The category includes four classes, from dry subhumid through semi-arid and arid to hyper-arid, so it covers far more than deserts: rangelands, steppe, savanna, and dryland forest are all inside it. Drylands make up 40.6% of global land area excluding Antarctica and are home to roughly 2.3 billion people, per the UNCCD's 2024 aridity assessment.
is a desert just unused land?
No. "Unused" describes the accounting, not the acre. A desert with intact cover is holding soil out of the air, routing and recharging water, carrying species that live nowhere else, and running cooler than the same ground stripped and graded — services that show up on someone's budget only after they stop. The perception of emptiness is partly a legacy of frontier land law, which defined desert land as ground that would not produce a crop without irrigation and set policy accordingly.
what to read next
The two spokes that take the two most common conversions:
- you don't have a wasteland. you have a living dryland — the dust, the dry spring, and the heat bill as receipts.
- greening the desert is not restoring a desert — the honest version of the tree-planting question.
Three things to look at:
- explore — places, agents, and what is currently funded.
- natural assets — the binder of what is being protected.
deserts.ensurance— the desert door as it actually stands today.
And one sibling argument from a different biome with the same structure: what grasslands actually are. Prairie gets called empty for the same reason and by the same reflex. It is a different stock with a different clock, and the misreading is identical.