Chemical weathering is the reason your garden isn't just crushed granite
Rocks sit there. Water hits them. Over thousands of years they become soil. That's the basic idea everyone gets from a fifth-grade science book. The actual mechanics are messier than that diagram suggests, and the part most people skip is that chemical weathering doesn't just break rocks down — it rearranges their entire molecular structure in ways that determine what grows on top of them for the next ten thousand years. I ran a soil composition study back in 2019 on a decomposing basalt outcrop in central Oregon. The surface looked like dirt after about six inches. Dig down two feet and you hit something more like coarse gravel held together by iron oxides. The transition zone between fully weathered material and fresh bedrock is where the actual chemistry happens, and it's not uniform. It patches. You get lenses of clay, pockets of remaining feldspar, zones where water tables have shifted over centuries and left behind secondary mineral deposits. If you're trying to model this process or predict soil behavior on a given site, treating it as a simple gradient from rock to dirt will give you wrong answers. The primary mechanism is hydrolysis. Water molecules, especially when slightly acidic from dissolved CO2 forming carbonic acid, attack the silicate bonds in minerals like feldspar and mica. These minerals break apart and recombine into clay minerals — kaolinite, illite, smectite — which have completely different physical properties than the original rock. Quartz doesn't react much at all. That's why sandy soils exist in places where the parent rock was rich in quartz but poor in feldspar. The quartz just physically breaks down over time while the rest of the mineral gets chemically consumed.
Oxidation is the other big one. Iron-bearing minerals like olivine and pyroxene react with oxygen and water to form iron oxides and hydroxides. Goethite and hematite. That's what gives red soils their color. In tropical environments with high rainfall and consistent warmth, oxidation and hydrolysis run fast enough that you can develop thick lateritic crusts in a few thousand years. In dry or cold climates, the same process takes tens of thousands of years and produces something closer to a thin regolith over bedrock. Carbonation affects carbonate rocks directly. Limestone and dolostone dissolve in weak carbonic acid. This is how karst topography forms — sinkholes, caves, underground drainage systems. The soil produced from limestone weathering is usually calcium-rich and tends to be less acidic than soil from silicate parent material, assuming everything else is equal. Which it never is. Here's the counter-intuitive part that trips up people who only read the textbook version: chemical weathering doesn't always produce more soil. In some conditions, particularly where leaching is intense and rainfall is heavy, the weathering process removes so many soluble cations — calcium, magnesium, sodium, potassium — that what's left is mostly silica and aluminum oxides. The volume actually decreases. You end up with a residue that's lower in total mass than the parent rock you started with. I've seen maps where the modeled soil depth didn't match the observed depth by a factor of three because the model assumed additive weathering when the site was actually experiencing net solute loss.
Another thing beginners miss is that biological activity isn't separate from chemical weathering. It's part of it. Plant roots exude organic acids — carboxylates, oxalates, citric acid — that chelate metal ions and pull them out of mineral structures. Mycorrhizal fungi do the same thing on a microscopic scale. The weathering rate in a forested watershed is orders of magnitude higher than in bare rock because the organisms are actively participating in the chemistry. You can't model soil formation accurately without accounting for the biological component, even if you're only interested in the abiotic reactions. The timescale is the hardest variable to pin down. Basalt weathers to soil in maybe two to five thousand years under temperate conditions. Granite takes longer because quartz is so resistant — on the order of ten to fifty thousand years for a meaningful profile to develop. Limestone dissolves faster than either but the "soil" it produces is thin and patchy because most of the mass goes into solution rather than staying behind as solid residue. Climate dominates all of this. Temperature and precipitation are the main controls, with biological activity as a significant modifier. If you're working with this practically — say you're evaluating a site for agriculture, construction, or remediation — the useful approach is to identify the parent rock, map the drainage and water table behavior, and sample the transition zone between weathered material and fresh bedrock. That transition zone tells you more than the surface layer ever will. The surface can be blown in, washed in, or deposited by glaciers and tell you nothing about the in-situ weathering that's actually happening. I learned that the hard way on a project where the topsoil looked fine for farming but the subsoil was still partially cemented volcanic tuff. Crops failed in year two when the roots hit the hardpan. The surface sample had lied.
Get the Full Details
Common mistakes in assessing weathering profiles include assuming uniformity across a site, confusing physical weathering (freeze-thaw, thermal expansion, pressure release) with chemical weathering, and not accounting for the difference between residual soil and transported soil. A lot of what looks like weathered material on a slope is actually colluvium — stuff that fell or washed down from somewhere else. It's not product of local weathering. It's product of gravity and water moving existing soil around. The tools you'd use in the field are basic. A soil pit or auger to examine the profile. pH testing of different layers. X-ray diffraction if you need mineral identification beyond what you can see. Sieve analysis for texture. None of this is expensive. The expensive part is doing it systematically across a representative area instead of grabbing one sample and calling it done. Chemical weathering is slow enough that you won't see it happen in real time, but fast enough that it's actively shaping the ground beneath your feet right now. The rate depends on rock type, climate, topography, biology, and time. Change any of those variables and the whole system shifts. There's no single equation that covers all of it, which is why field work still matters despite all the modeling software available.
