

But in either case the seepage would probably be less than if I had breached the actual river involved, with its gravity and (for a river) continuing feed.) (I've only made this error long ago, when there were just normal aquifers, no weaker ones.

The worst case scenario (not sure how common to encounter a boundary in just the right/wrong spot) is if you have an actual aquifer inhabiting the layer immediately passing under a clump of water (river, pond, pool, ocean) and you dig into it knowing that the warning+cancel was just the liquid tile (from checking what you can see of that), expecting a sufficient floor/ceiling separation from actual-wetness, but then discover the rock-seep exists and have to deal with the consequences of that. It is more logical with detecting magma (you can feel heat through that layer, and also sideways through walls, without expectation of 'seeping', though if you overstep the mark and (wrongly) ignore the warming-warning you've got a potentially larger issue on your hands, and feet. Yes, inconsistent in that you "detect dampness" beneath the river-bed (you can assume the 'floor' of that water-tile is impervious and not waterlogged, despite the dampness-detection) but an aquifer tile is wet for the whole rocky volume, including the " floor slice", so digging below exposes you to the water in the wet upper.
