Nobody gets new water

Start with the physical claim, because that part is checkable.

Earth holds something near 1.386 billion cubic kilometers of water, and has held about that much for longer than there have been people to measure it. The water cycle manufactures none of it. It relocates what already exists: ocean to air, air to land, land back to ocean, with long detours through ice, rock, and living things. Every glass of water you have ever drunk has been through all of that, repeatedly.

The leaks are real and they are small. Sunlight splits water vapor high in the atmosphere, and the loose hydrogen escapes to space at roughly 95,000 metric tons a year. That same process is what cost Venus its oceans. Against Earth's total it is a rounding error. Volcanoes bring mantle water up. Subduction carries surface water down, and there may be several oceans' worth locked into minerals far below anything anyone has drilled. Whether the surface budget is gaining or losing across hundreds of millions of years is still argued about by people who do it for a living.

On any timescale a person lives inside, the amount is fixed. None of that is controversial. It is what hydrology teaches on the first day.

Fixed and available are different words

Here is where the fixed-total fact gets misused, including by people who are fond of it.

Ninety-seven percent of that water is salt. Another two percent is locked in ice. Contamination takes water out of use without taking a single molecule off the planet. And then there is the case that should trouble anyone who has looked at it: the Ogallala under the American plains, the aquifer under the Arabian peninsula, the one under the North China plain. Fossil water, accumulated across tens of thousands of years, being pumped out across decades.

Nothing is destroyed when that happens. The water moves from a store that recharges over millennia into the fast cycle that recharges over weeks. On a human horizon that transfer runs one way. The planetary ledger balances perfectly and the well still goes dry.

So the quantity is conserved and the availability is not. Both things are true, and only one of them is comforting.

Which makes distribution the entire question

If the total cannot change, then everything that matters about water is where it sits, when it arrives, what form it takes, and whether anybody is standing nearby.

Weather is what decides all four of them, season by season and region by region.

That is the part of this I hold as more than hydrology. A closed system with a fixed quantity, whose distribution is adjustable region by region and season by season, is a remarkably efficient instrument. It requires no new creation to bless a valley or to leave one wanting. The same water, routed differently, does both.

The reading is older than I am

I did not come up with this, and the texts are unusually direct about it.

Deuteronomy 11 makes rain conditional in plain terms, promising it in due season and warning of a heaven shut up. Elijah opens his account in 1 Kings 17 by announcing a drought rather than praying for one. And Amos 4:7 states the regional version almost word for word: rain caused to fall on one city and withheld from another, one field watered while the next withered.

That last one is the verse I would point at. It is not describing weather as indifferent machinery that God occasionally interrupts. It describes differentiation as the point -- two fields, one rain, deliberately.

A category I had been missing

Writing this down showed me a gap in my own frame, which is the usual reward for writing something down.

Weather is one mechanism. Storage is another, and it works differently.

An aquifer is not rain arriving or being withheld. It is rain that already arrived, across tens of thousands of years, and was kept. A blessing given and held in reserve rather than one granted or denied in the moment. Which means pumping it dry inside a single lifetime is not an instance of blessing withheld at all. It is an instance of blessing received and spent.

Those call for different responses, and lumping them together lets the second one hide inside the first. A region can be short of water because rain did not come. It can also be short because it drank an ice age's worth in two generations and called the consequence a drought.

The part I will not claim

The inference does not run backward, and I want to say so before anybody else says it for me.

If drought is sometimes an instrument, it does not follow that every dry place is a guilty one. The tradition closes that door itself and does not leave it to me. Matthew has the rain falling on the just and the unjust in the same sentence, which wrecks any tidy ledger you might want to build. And when the disciples try exactly this inference in John 9 -- affliction implies someone sinned, either him or his parents -- they are told flatly that it does not.

So I hold that weather is an instrument, and I do not hold that I can read any particular dry season backward to a verdict. Anyone who tells you they can is selling something. The reading I am describing explains how a blessing could be delivered or withheld. It does not license me to look at a map of water stress and start assigning blame, and the map would not support me if I tried: the driest and most desperate places are mostly the poorest, which is the opposite of what a simple ledger would predict.

The flood may not be the exception I thought

For a long time I carved out the flood as the one event where the fixed-quantity rule broke, because that much water had to come from somewhere.

Reading Genesis again, I am not sure it needs carving out. Genesis does not describe rain alone. The fountains of the great deep break open, which is water arriving from a subterranean store rather than water newly made. That is relocation at catastrophic scale -- the same mechanism as every ordinary storm, run at a magnitude nobody has seen since.

Which would mean the rule held even then. The same water, moved.