Runoff from dry ground
A Socratic walk-through of runoff from dry ground — reasoned out one step at a time, not lectured.
The question we started with
THE QUESTION #Why does a downpour on parched ground mostly run away instead of soaking in?
A drought breaks. The first heavy storm arrives on ground that has been waiting for it for months, and instead of vanishing into the soil the water sheets off, gathers in the gullies and floods the town downstream. The soil was as empty as it ever gets, and it took almost none of it. That reads as a contradiction: an empty sponge should be thirstier than a full one. So which part of the sponge picture is wrong?
Reasoning it through
REASONING #Start by separating two quantities we habitually confuse. One is capacity — how much water the soil could eventually hold. The other is rate — how fast it can accept water at the surface, which hydrologists call infiltration capacity. Only the second decides what happens during a storm. If rain falls faster than the soil can take it in, the surplus stands on the surface and then flows, no matter how much room lies below. Drought empties the reservoir; it says nothing directly about the width of the inlet.
So the question becomes: what could shrink the inlet on dry ground? Two mechanisms, and both are physical and checkable.
The first is that the surface can stop being wettable at all. Soils accumulate organic compounds from roots, fungi and decaying litter, and some of these are waxy and water-repellent. While the soil is moist they are held apart by water films and cause little trouble. Let the soil dry past a threshold and they coat the particles directly, and the ground behaves like a waxed jacket: a drop sits on it as a bead rather than sinking. Fire makes this dramatically worse — burning volatilises organic compounds which then condense a little below the surface, leaving a repellent layer over the top of a burnt slope. This is why the season after a wildfire is the season of debris flows.
The second is that intense rain builds its own lid. A large raindrop lands with real energy. On bare soil it shatters the crumb structure at the surface, and the freed fine particles wash into the pores just below and clog them. Within minutes there is a thin dense seal — a crust — perhaps a millimetre or two thick, and infiltration through it can fall by an order of magnitude. Notice what makes this vicious: the same storm intensity that demands a high infiltration rate is what destroys it, and the destruction is fastest on ground with no vegetation to break the fall of the drops. A drought that has killed the cover has arranged exactly those conditions.
Now push back on our own explanation, because a good part of the folk version is wrong. Is dry soil generally bad at taking water? No — the opposite, usually. Dry soil pulls water in by capillary suction far more strongly than wet soil, so initial infiltration into dry ground is typically high and then declines as the profile fills. Dry clay soils crack open as they shrink, and the first rain pours straight down the cracks. Dryness by itself does not cause runoff. What causes it is repellency, or crusting, or both — and dryness is a condition that makes those two more likely, not a cause in its own right.
The analogy
THE ANALOGY #Think of a funnel with a fine mesh across its throat. How much the jug beneath can hold is one thing; how fast the mesh passes water is another, and only the second matters while you are pouring. Grease the mesh and water beads on top of it. Pour hard enough to wash silt into it and it blinds itself. Either way the jug beneath is still empty, and still not filling.
a funnel's mesh is fixed and passive, whereas soil surface is alive and continually rebuilt — roots, worms and fungi open new channels, freeze-thaw and wetting cycles break crusts apart, and repellency itself relaxes once the soil has been wet long enough. The throat is repaired as fast as, or faster than, it is blocked, which is why the same field can behave completely differently in two consecutive years.
Clarifying the model
THE MODEL #Two refinements. First, what we have described is only one of the ways runoff is generated. This kind — rainfall arriving faster than the surface can accept it — dominates in dry, bare and crusted landscapes. In a wet temperate catchment the more usual mechanism is the opposite: the soil is already saturated from below and simply has nowhere left to put anything. Same flood, entirely different cause, and the remedies differ accordingly.
Second, and this is the honest correction to the whole framing: dryness is not the main variable. Cover and structure are. Ground held by vegetation and litter absorbs the raindrops' impact so no crust forms; roots and soil fauna maintain the large pores that carry water down quickly; stable aggregates resist slaking. A well-covered soil in a drought will still take a downpour. A bare, compacted, low-organic-matter soil will shed it whether it is dry or not. Drought is the accelerant here, not the fire.
A picture of it
THE PICTURE #How to readR1 is the outcome, and the arrow out of it points at the one condition that actually produces it — C1, rain outrunning the surface. C2 and C3 refine that condition by naming the two ways a dry surface shrinks its own intake, and C3 in turn depends on C4, bare ground. The two elements below are landscapes: a burnt hillslope satisfies both the repellency and the bare-ground conditions, which is why it floods so readily, while a dry but well-covered pasture satisfies only the intensity condition — and the arrows that are missing from it are the point, since without crusting or repellency the rain still goes in.
What became clearer
WHAT CLEARED #The sponge picture fails because it measures the wrong thing: a storm is decided by the rate at which the surface accepts water, not by the room available underneath. Dry ground can shrink that rate two ways — organic coatings that make it refuse to wet, and a crust that intense rain builds for itself out of the soil's own broken crumbs. But dryness is a risk factor, not the mechanism. The variable that really decides whether a drought ends in a flood is whether anything is still growing on the surface when the rain arrives.
Where to go next
ONWARD #- Why the first storm after a wildfire produces debris flows out of all proportion to its size.
- How saturation-excess flooding in wet catchments calls for the opposite management to this.
Key terms
TERMS #| Term | What it means |
|---|---|
| Infiltration capacity | the maximum rate at which a soil surface can take in water, distinct from how much it can hold. |
| Soil water repellency | the state in which hydrophobic organic coatings on dried particles make soil resist wetting. |
| Surface seal or crust | a thin dense layer formed when raindrop impact breaks aggregates and fines block the pores beneath. |
| Infiltration-excess overland flow | runoff generated because rain outpaces the surface, as opposed to because the soil is already full. |
Every term the collection defines is gathered in the glossary.