Drought, Heavy Rain, and Foundation Movement

Regional rainfall usually gets quoted as an annual figure, which is the least useful form it comes in. The National Weather Service puts the normal precipitation at Dallas/Fort Worth at 4.78 inches for May and 2.08 inches for July, against an annual normal of 37.01 inches[1] — more than twice as much rain in a normal May as in a normal July, in a record that has been kept at that station since September 1898[1].

And those are the normals, which are averages over a great many years. Any individual year departs from them in both directions, sometimes severely. The soil does not respond to the annual figure at all; it responds to the sequence. A year that lands exactly on the annual normal can still contain a rainless stretch through the worst of the heat followed by a month that makes the total back up in a fortnight.

That alternation is the engine. A climate delivering the same rainfall smoothly across twelve months would move foundations far less than one delivering the identical amount in bursts separated by drought.

As expansive clay loses water it loses volume, and here it does so visibly. The USDA's description of Houston Black records cracks 0.5 to 4 inches wide extending from the surface to a depth of 12 inches or more, remaining open for 90 to 150 cumulative days in most years[2]. Three to five months a year of open ground cracks is the ordinary condition of this soil, not an emergency.

The part that matters for a house is that the ground does not dry out evenly. Soil at the perimeter is exposed on one side to sun and wind and to whatever trees and shrubs are drawing on it, while soil under the middle of a slab is shielded by the slab itself. Drying therefore starts at the edges and works inward, which means the support under the outer few feet of a foundation can be shrinking while the support under the centre has barely changed.

That is why a long dry summer produces the classic late-season complaints — a door that has started to bind, a crack over a window that was not there in the spring. The edge has moved relative to the middle.

When a dry spell finally breaks, two things happen to that soil at once, and they pull in opposite directions.

First, the clay is poor at absorbing rain quickly. The same USDA description gives the series a permeability class of very slow, with runoff rated high to very high depending on slope[2], so a substantial part of a heavy rain leaves across the surface instead of soaking in where it fell. Second, and cutting the other way, the cracks that opened during the drought are an open path carrying water straight down past the surface to depth.

The combination re-wets the perimeter quickly and unevenly — fastest where the cracking was worst and where the runoff collects, slowest under the interior. The soil swells back, and the parts of the house that dropped over the summer come back up, not necessarily in step with one another.

This is the honest explanation for something that confuses a lot of homeowners: a door that stopped closing in September works fine again by March. Nothing was repaired. The soil went back the other way. That is a description of movement, not of its absence, and the same cycle will run again.

If symptoms appeared in a particular season, it is worth knowing what that season actually was rather than relying on memory. The US Drought Monitor is the standard record for this. It is a map released every Thursday, classifying conditions from D0 (abnormally dry) through D1, D2, D3 and D4 (exceptional drought)[3], and it is built from precipitation, streamflow, reservoir levels, temperature and evaporative demand, soil moisture and vegetation health, with input from a network of more than 450 experts[3].

Because it is archived weekly, you can look up the category your county sat in during the month a symptom appeared, and again during the month it eased. Put that next to your own dated notes and photographs and you have something a professional can use: not "it got worse last year" but a symptom log with the regional soil-moisture conditions attached to each entry.

Pair it with the perimeter observations on the drainage page, which covers the local half of the same question. The regional record tells you what the whole county experienced; your own downspouts tell you what one wall did with it.

The Drought Monitor is explicitly not a forecast — it looks back in time, providing a snapshot of very recent conditions[3]. It describes what the region went through. It says nothing about what any individual slab did in response.

A severe drought year is context. It is not a diagnosis, and it does not establish that a particular house was damaged by it — most houses in the same conditions are not.

It works in the other direction too. A symptom that receded when the rain came is not proof that nothing is wrong, because seasonal recovery is exactly what early movement on expansive clay looks like. What separates a house riding the cycle from a house progressively losing support is what the same measurements say across several cycles — which is an argument for writing them down now rather than waiting.

If something is already showing, the symptom check will weigh it alongside whatever else you have noticed, and the soil mechanism page covers why the ground behaves this way in the first place.