Geothermal Gradient: The Real Number, and Why the Average Misleads

How fast does the Earth get hotter as you dig down? The most fully documented US measurement comes from the USGS, whose 1987 compilation of 284 temperature-gradient measurements across the conterminous United States found an average of 29 °C per kilometer of depth. That average needs context: the same dataset carries a standard deviation of 11 °C/km, and the individual measurements run from 6 all the way to 69 °C/km.

The spread matters as much as the mean. The average is genuine — measured, published, reproducible — and on its own it is not enough to predict any single spot on the map. At the extremes of that dataset, one site warms about eleven times more slowly than another. A useful answer therefore needs the mean, its source, and the conditions under which it applies.

What Is the Geothermal Gradient?

Drill a hole and lower a thermometer, and the reading climbs as you descend. The USGS defines it plainly: "This increase in temperature with depth is called the geothermal gradient."

The units tell you what kind of quantity this is: degrees Celsius per kilometer (or, if you prefer, degrees Fahrenheit per 100 feet). It's a rate — how steeply temperature rises with depth — not a temperature, and not an amount of energy. That distinction sounds pedantic until you watch people mix the gradient up with heat flow, which is measured in milliwatts per square meter and is a genuinely different physical quantity. We'll separate the two properly below, because conflating them is a common error in explanations of geothermal gradients.

The gradient is why geothermal energy is reachable at all: the deeper you go, the hotter the rock, so with enough drilling you can reach temperatures useful for anything from district heating to electricity generation. Whether a given site is a usable resource depends on more than temperature — permeability, fluid, and drilling economics all matter — but the gradient is the precondition.

The Measured Number: 29 °C/km — and a Huge Spread

The dataset behind this article is USGS Open-File Report 87-592, a compilation of geothermal-gradient data in the conterminous United States. In the authors' words: "The average value of all 284 gradients for the United States is 29°C/km. The average for the eastern United States is 25°C/km and for the west is 34°C/km." And on the spread: "The gradients lie between 6 and 69°C/km... The mean of the 284 gradients is 29°C/km with a standard deviation of 11°C/km."

Broken out regionally:

The western mean is 34 °C/km against the East's 25 — the West runs hotter, though both regions carry a standard deviation of about 10–11 °C/km, so neither is tightly grouped. That nine-degree-per-kilometer difference between regional means is not a rounding error; over a 3 km well it's roughly 27 °C of rock temperature.

Three scope notes that go with these numbers wherever they travel:

1. This is a conterminous-United-States dataset, compiled in 1987. It is not a global average, not an "Earth's crust" average, and not a constant of nature. It's what 284 US drill holes showed.
2. The measurements come from drill holes generally deeper than 600 m, and the report treats them as regionally applicable to about 2 km depth. Below that, extrapolate at your own risk.
3. Convective hydrothermal systems were excluded. Places where hot water is actively moving through the rock aren't in this average, which describes conductive background conditions.

Local conditions determine whether the national figure is informative for a particular borehole. For a specific site, a local measurement is more useful than the national mean — and obtaining one is what site assessments and test bores are for.

Where Does "25–30 °C per km" Come From?

Search this topic and you'll find the same phrase everywhere: the geothermal gradient is "about 25–30 °C per kilometer." It is widely repeated in general explainers of the subject, and it is usually presented without saying whose measurements it comes from or what area they cover.

No authoritative publication located for this article defines 25–30 °C/km as a global, continental-crust, or upper-crust average with a stated measurement population behind it. The interval gets repeated from page to page, but the sources reviewed here do not say whose measurements it summarizes, over what area, at what depths.

The figure is plausible, but its provenance is unclear. The USGS eastern-US mean is 25 °C/km and the all-US mean of those 284 measurements is 29 °C/km (standard deviation 11, range 6–69), so "25–30" brackets two legitimate regional means. The source of the bare interval remains unidentified, as does the population it is meant to describe. It may well trace back to legitimate published work, but the pages that repeat it do not cite one.

The problem isn't that 25–30 is wildly wrong. The problem is what the phrasing hides: stated as a bare interval, it reads like a physical constant with a ±10% tolerance, when the measured reality is a mean with a standard deviation of 11 °C/km and an observed range spanning roughly an order of magnitude (6 to 69). "The gradient is 25–30 °C/km" and "US gradients average 29 °C/km but run from 6 to 69" lead a reader to very different decisions — about drilling, about siting, about what's under their own feet.

Gradient Is Not Heat Flow

Two quantities get tangled together in many discussions of Earth's heat, and they need untangling.

The geothermal gradient is a rate of temperature change with depth: °C per km. Heat flow is a flux of energy: milliwatts crossing each square meter of surface, mW/m². They are related but not interchangeable, and the thing linking them is thermal conductivity — how readily a given rock conducts heat.

High-conductivity rock carries a given heat flow toward the surface with a smaller temperature gradient; low-conductivity rock needs a steeper one to move the same heat. Same heat flow, very different gradients, depending on what the rock is made of.

This is why maps of the two quantities don't line up. The USGS says so directly: "Patterns of temperature gradients are similar to those for heat flow in some areas, but there are significant differences caused by regional differences in thermal conductivities." A region can have unremarkable heat flow but a steep gradient because its rocks conduct poorly, or vice versa. One map cannot be read off the other without thermal-conductivity data.

For scale, the measured heat-flow numbers come from Davies & Davies (2010): mean continental heat flow is 70.9 mW/m², mean oceanic heat flow is 105.4 mW/m², and the global mean is 91.6 mW/m². Notice the units — those are energy fluxes, and none of them can be converted into a "°C per km" figure without knowing the conductivity of the rock in question. Using one quantity as though it were the other is physically incorrect.

Where the Heat Comes From

Why is there a gradient at all? Because the Earth's interior is hot and the surface is cold, and heat flows down that temperature difference toward space.

Two well-sourced numbers frame the budget. Davies & Davies, integrating heat-flow measurements over the whole planet, conclude: "Our final preferred estimate is 47±2 TW" for Earth's total surface heat loss. On the supply side, McDonough and colleagues put radioactive decay's contribution directly: "Earth's radiogenic heat budget is 20 TW" — heat generated continuously by the decay of radioactive elements in the Earth's interior.

Divide those two figures — our own arithmetic, not a published partition — and radiogenic decay accounts for roughly 43% of the heat the Earth loses at its surface. The remainder is heat the planet is losing faster than radioactivity replaces it; the cited sources measure the totals but do not break that residual down further, and it is not broken down here.

The practical upshot: a substantial share of the heat leaving Earth is continuously generated by radioactive decay, and the total is large compared with human energy use. That's the physical basis for treating geothermal as a renewable resource, a question we take up properly in whether geothermal energy is truly renewable.

Why Your Ground Loop Uses Shallow Ground Temperature, Not the Deep Gradient

For homeowners, the key distinction is between the deep geothermal gradient and the shallow ground temperature a heat pump actually uses — two quantities that are often confused.

A reasonable homeowner hears "the Earth gets hotter as you go down" and concludes that a geothermal heat pump works by tapping that rising heat. It doesn't. Using the USGS mean of 29 °C/km (from those 284 US measurements, standard deviation 11, range 6–69) — our own arithmetic — the gradient's contribution is:

• Over 30 feet — the depth DOE cites for stable ground temperatures — the gradient adds about 0.27 °C, or roughly 0.5 °F.
• Over 400 feet — a deep residential borehole — it adds about 3.5 °C, or roughly 6.4 °F.

Half a degree Fahrenheit at 30 feet. Even a deep vertical bore picks up only a handful of degrees from the gradient, and at a site on the low end of the measured range it would be a fraction of that. The gradient is real, but at residential depths its contribution is small next to the shallow ground temperature the loop actually uses.

The DOE puts numbers on that shallow ground temperature: "Temperatures at about 30 feet below the surface remain relatively constant year-round—between about 50°F (10°C) and 59°F (15°C)." Geothermal heat pumps, in DOE's words, "use the constant underground temperatures of the shallow earth as thermal storage that enables efficient heating and cooling." The shallow ground sits near the local mean annual surface temperature and holds it steadily — which is why DOE can give a single 50–59 °F band for shallow ground across most of the country while deep gradients vary widely beneath it. In winter that steady ground is far warmer than the outdoor air; in summer, far cooler. A heat pump exploits that seasonal temperature difference, not the gradient; the full mechanism is covered in how geothermal heating works.

The conclusion worth carrying away: a backyard ground loop is not tapping the geothermal gradient. It's using stable shallow ground as a thermal battery. This is also why residential loops don't need to chase depth for its own sake — as our geothermal well depth guide covers, bore depth is sized for heat-exchange capacity, not for reaching hotter rock. If someone tries to sell you a deeper, pricier borehole on the promise of "getting into the geothermal gradient," the numbers above are your defense.

What the Gradient Means for Geothermal Power

Where the gradient genuinely earns its keep is geothermal power — the utility-scale plants that drill kilometers, not hundreds of feet.

Power generation needs hot fluid. DOE's Geothermal FAQs put the bar here: "Geothermal fluid temperature should be at least 300°F/149°C, although plants can operate on fluid temperatures as low as 210°F/99°C." Note carefully what kind of number that is — a resource temperature threshold, not a gradient threshold. No authoritative source identified for this article publishes a "favorable gradient" cutoff for power siting, so none is offered here.

But the reasoning connecting gradient to siting is straightforward. If the target is a given fluid temperature, a steeper local gradient means you reach that temperature at shallower depth — and since drilling is a dominant cost, shallower means cheaper. A site warming at the top of the USGS-observed range hits useful temperatures at a fraction of the depth needed at a site near the bottom of it. That's why the western US, with its higher mean gradient (34 °C/km across 147 USGS measurements) — and why regions with steep local gradients generally — attract power development, and it's why real projects start with measured local data rather than any national average. The drilling side of that equation, including what those deep wells involve, is covered in geothermal drilling and wells.

One caveat from the dataset itself: the USGS gradients are regionally applicable to about 2 km and exclude convective hydrothermal systems. Power prospects often involve exactly the depths and the convective settings that compilation set aside — one more reason the 29 °C/km mean informs intuition but never substitutes for site measurement.

Converting the Units

US drilling talks in feet and Fahrenheit; the literature talks in kilometers and Celsius. These conversions are our own arithmetic:

So at the US mean, every 100 feet of depth buys you about a degree and a half Fahrenheit. It also makes the heat-pump arithmetic above easy to check against any quoted borehole depth.

Frequently Asked Questions

What is the geothermal gradient?

It's the rate at which temperature increases with depth below the Earth's surface, measured in °C per kilometer (or °F per 100 feet). The USGS definition: "This increase in temperature with depth is called the geothermal gradient." In the USGS compilation of 284 US measurements, gradients averaged 29 °C/km with a standard deviation of 11 and a range of 6 to 69 °C/km.

Is the geothermal gradient 25–30 °C per km?

That interval is quoted everywhere, but no authoritative source identified for this article defines it as a global or crustal average with a stated measurement population. What is actually published: USGS measured 284 gradients across the conterminous US and found a mean of 29 °C/km (eastern mean 25, western mean 34), with a standard deviation of 11 and individual sites ranging from 6 to 69 °C/km. So 25–30 brackets real US regional means — but any specific location can fall far outside it.

What's the difference between geothermal gradient and heat flow?

Gradient is a rate of temperature change with depth (°C/km); heat flow is energy crossing a unit of surface area (mW/m²). Rock conductivity links them — conductive rock carries the same heat with a gentler gradient — so the two don't map onto each other one-to-one. USGS notes their patterns differ regionally precisely because of conductivity differences.

Does the geothermal gradient matter for my home's geothermal system?

Practically, no. At the USGS mean of 29 °C/km (across 284 US measurements, SD 11, range 6–69), the gradient adds only about 0.5 °F over 30 feet and about 6.4 °F over a 400-foot borehole — our arithmetic. Residential systems work because shallow ground holds a steady 50–59 °F year-round (DOE), warmer than winter air and cooler than summer air. The gradient is a factor for deep geothermal power drilling, not backyard loops.