The Dirt Under the Dirt

Column #30: The Dirt Under the Dirt

Every Saturday night somebody walks back to the trailer and says "the track changed on me." No. The track did not change on you. You never knew what the track was. You were reading the top eighth-inch of a system that goes down four feet. The surface you drove on tonight is layer three of a three-layer stack, and if you do not understand layers one and two, you are guessing. You have been guessing. That ends today.

Wikipedia will tell you dirt track racing happens on clay. It will not tell you that the native soil under Eldora Speedway is a Celina silt loam — an Aquoll with a saturated hydraulic conductivity of 4.2 µm/sec and a drainage class of "somewhat poorly drained." It will not tell you that this is why Eldora holds moisture three hours longer than a track 90 miles east sitting on Udipsamments — sandy, excessively drained soil that bleeds water like a screen door on a submarine. It will not connect any of this to your right rear tire pressure. I will.

The Three-Layer Model

Forty years of reading dirt and the single most useful framework I ever built fits on a napkin. Three layers. Three different data sources. Three different timescales. Every setup mistake I have ever seen — and I have seen thousands — comes from confusing which layer the driver is reacting to.

Layer 1: Substrate. The native soil under the track. It was there before the bulldozer showed up. It will be there after the promoter retires. It controls drainage, moisture retention, and the baseline behavior of everything above it. Timescale: geological. You cannot change it. You can only understand it.

Layer 2: Racing Surface. What the promoter imported, blended, and maintains. Clay from a pit 40 miles away. Calcium chloride for moisture retention. Sawdust for cushion structure. This is the recipe, and it sits on top of the substrate like frosting on a cake. Timescale: seasonal to annual. The promoter changes it. You adapt to it.

Layer 3: Race-Night Condition. What is on top right now. Tack, slick, rubber, moisture gradient. This is what your track walk tells you, what your eyes read during hot laps, what your tires feel in turn one. Timescale: hours. It changes between your heat and the feature. This is the layer most racers think is "the track." It is not the track. It is tonight's weather report on a surface built on a substrate you have never looked up.

Layer 1: What SSURGO Actually Tells You

The USDA Natural Resources Conservation Service has mapped the soil of the United States at a resolution that would make a race team's data acquisition system jealous. Every square meter has a map unit key — a mukey — tied to a taxonomic classification, particle-size distribution, drainage class, available water capacity, and saturated hydraulic conductivity. The system is called SSURGO. The farmer interface is Web Soil Survey. The API — the one our dataset uses — is Soil Data Access at SDMDataAccess.sc.egov.usda.gov. Same data agronomists use to decide whether to plant corn or soybeans. We use it to understand why your 410 goes slick in 14 laps at one track and holds moisture for 30 at another.

We have measured substrate data on 557 of 598 tracks in our dataset. That is 93%. The remaining 42 return no data at the pin — usually because the geocode lands on an urban land map unit (the fairgrounds is at the edge of town but the coordinate points at the courthouse downtown) or the track is international and ISRIC SoilGrids hit a rate limit. Fix path is straightforward: correct the geocode, re-run the query, or use a 2 km buffer ring and take the dominant soil in the surrounding area. We mark those INFERRED and never present them as measured. Honesty about your data is as important as the data itself.

The numbers that matter for racers are these:

Sand / Silt / Clay percentage. The horizon-zero texture — the top layer of native soil. A track sitting on 85% sand, 8% silt, 7% clay (Udipsamments, common in parts of Florida and the Sandhills) behaves radically differently from one sitting on 22% sand, 45% silt, 33% clay (Aquolls, common in Ohio and Indiana river bottoms). The sand-dominant substrate drains water almost instantly. The clay-silt substrate holds it for hours.

Drainage class. SSURGO assigns seven classes from "excessively drained" to "very poorly drained." Excessively drained substrates lose moisture fast — the track goes slick early, often by the B-main. Poorly drained substrates hold water, which means the promoter fights saturation instead of dryness, and the track can stay heavy and tacky deep into the feature if weather cooperates.

Available Water Capacity (AWC). Measured in inches of water per inch of soil. This is how much moisture the native soil can hold in the plant-available range — the range where it is neither bone-dry nor saturated. For racing, AWC tells you the substrate's buffering capacity. High AWC (0.18–0.22 in/in, typical Aquolls and Udalfs) means the substrate acts like a sponge under the racing surface, feeding moisture upward through capillary action even after the surface dries. Low AWC (0.04–0.08 in/in, sandy Psamments) means once the surface water is gone, there is no reservoir underneath. The track dies fast.

Saturated Hydraulic Conductivity (Ksat). Measured in µm/sec. This is how fast water moves through fully saturated soil. A Ksat of 40+ µm/sec (sandy substrate) means rain drains through fast — good for avoiding cancellation, bad for surface moisture retention. A Ksat under 4 µm/sec (clay-heavy substrate) means water pools and perches, the track stays wet longer, and after a heavy rain the promoter may need 4–6 extra hours to get the surface right.

Taxonomic suborder. This is the classification that ties it all together. Aquolls are wet prairie soils — poorly drained, high clay, high organic matter, high AWC. They dominate the corn belt and Ohio valley. Udipsamments are sandy, well-to-excessively drained, low AWC — coastal plains, lake-bottom deposits, glacial outwash. Udalfs are the "good" forest soils — moderate everything, well-drained but not excessive. Aquepts are wet mineral soils without the organic content of Aquolls — they hold water but do not have the same capillary buffering. Each one creates a different baseline for the promoter to work with.

Layer 2: What the Promoter Builds

Here is the mistake that 90% of racers make — and the mistake that would trap you if you only looked at SSURGO data. The clay percentage at the SSURGO pin is the native soil. It is not what you are racing on. A track in central Florida might sit on 87% sand substrate. Pull up Web Soil Survey, look at the pin, and you would think the track could not possibly support a dirt-racing surface. But the promoter trucked in 200 loads of Hawthorne clay from a pit in Alachua County, blended it with the native sand, and built a racing surface that runs 31% clay in the top 6 inches. The SSURGO data says sand. Your tires say clay. Both are true. They are measuring different layers.

The target for most racing surfaces is 28–35% clay content in the amended layer. This range provides enough cohesion to hold moisture and rubber, enough silt for a smooth polished surface when it slicks off, and enough aggregate structure to resist breaking into dust too early. Below 25% clay, the surface sheds moisture and gets dusty. Above 38%, it gets heavy and sticky when wet, then turns to concrete-hard slick when it dries — a narrow window that makes the promoter's job miserable.

What promoters add:

Imported clay. The primary amendment. Most promoters have a preferred clay source — a specific pit, a specific blend, a specific color. Red clay from the Piedmont is iron-oxide-rich, binds well, and polishes to a predictable slick. Gray-blue clay from river-bottom deposits is higher in montmorillonite, swells more when wet, and holds moisture longer but can get gumbo-sticky. The promoter's clay source matters as much as the substrate underneath.

Calcium chloride (CaCl₂). Hygroscopic salt. Draws moisture from the air. Applied at 0.5–1.5 lbs per square yard, it helps sandy-substrate tracks retain surface moisture during the program. Tracks on Udipsamments — sand-dominant, excessively drained — are the heaviest CaCl₂ users. A track on Aquolls rarely needs it. The salt changes tire wear rates: more CaCl₂ means softer surface, more tire degradation, and a stickier feel in the first 5 laps that can trick you into thinking the track is tackier than it will stay.

Sawdust and organic amendments. Mixed into the surface layer for cushion structure. Sawdust creates voids that hold moisture and provide a "give" to the surface — the cushion that absorbs car movement instead of deflecting it. Too much sawdust and the surface gets punky, breaks apart in chunks, and creates a rough, unpredictable cushion. Too little and the surface polishes to a one-lane bottom track by lap 15.

Water management. Not an amendment in the chemical sense, but it is part of the promoter's recipe. How much water, how often, which corners get more — this is layer-2 infrastructure work that happens before the gates open. A track with buried irrigation lines in the substrate can feed moisture from below, maintaining surface consistency through a 40-lap feature. A track with just a water truck on top is fighting evaporation every minute.

Layer 3: What Your Track Walk Tells You Tonight

This is the layer most racers live in. It is the only layer you can read with your boots. And it is the layer that changes fastest — sometimes between your heat race and the feature, sometimes between turns one and three on the same lap.

Your track walk gives you five data points:

1. Footprint depth. Walk the racing groove. How far does your boot sink? Less than 1/8 inch: slick, polished, low grip. The track has already transitioned. 1/4 to 1/2 inch: tacky. The surface has moisture and structure. Tires will bite. More than 1/2 inch: heavy. Wet clay. The track has too much moisture — either freshly watered or poorly drained substrate feeding from below. Heavy tracks eat right rear tires and punish cars with too much stagger.

2. Surface color gradient. Dark means wet. Light means dry. Walk from the infield edge to the wall and note where the color changes. If the bottom 10 feet is dark and the top 15 feet is light, the bottom has moisture and the top does not. By lap 20, the bottom will still have grip and the top will be glass. If the entire surface is uniform medium-brown, moisture distribution is even — the promoter did good work.

3. Chunk test. Pick up a piece of surface material. Squeeze it. If it holds together in a ball and takes fingerprint impressions: tacky, good moisture, 20–25% moisture content. If it crumbles immediately: dry, less than 12% moisture, the track is already slick or heading there fast. If it oozes water between your fingers: too wet, above 30% moisture, the surface will be heavy for the first 10 laps at minimum.

4. Corner comparison. Walk all four corners. They will not be the same. Sun exposure dries turns 3 and 4 first at most tracks (western exposure catches afternoon sun). Wind accelerates evaporation on the windward side. A 15 mph southwest wind dries turns 3 and 4 at Route 66 Motor Speedway in Amarillo two hours faster than turns 1 and 2. The substrate is the same under all four corners, but tonight's condition is different in each one. Your setup has to handle both.

5. Rubber check. Look for dark, shiny streaks in the groove. That is rubber laid down by previous races — sometimes from the last event, sometimes built up over a season. Rubber on the surface changes grip characteristics more than anything the promoter does tonight. Rubber is a tire compound layer sitting on top of a clay layer sitting on top of native soil. Three layers plus one. If the bottom groove has heavy rubber and the top does not, the grip differential between bottom and top can be 30–40%. That is why the fast car on the bottom pulls away while the car on top slides like it is on ice.

Where the Layers Interact — Class by Class

The three-layer model is not academic. It drives setup decisions in every class. The physics differ because the cars differ.

Common Mistakes — And the Wrong Numbers People Use