How do rivers freeze?

My office looks out a window onto a fairly large river. Today there was surface ice on both banks, with clear water in the middle. As the day has progressed, the gap has “filled in” from the downriver direction, so that downriver there is now ice from bank to bank.

My question is, what exactly is the process for a river freezing? I imagine it starts at the banks because shallower water gets cold more quickly? Does the movement of the water have an effect? If I’m a drop of water flowing down the river, do I just kind of crystallize mid-flow as I drop below freezing, and maybe thaw again as I am turned under into the not-quite-freezing water below, until I happen to catch against the sub-zero ice from the shallower portions and stick?

It’s kind of fascinating to watch.

crystals tend to grow on crystals or other solids. so the water will freeze from the shore or banks outward.

the ice surface might start even a bit above the flowing water.

There’s the “heat of fusion”(?); to go from water at 0C to ice at 0C the water must lose 80 calories. This is a lot easier if it is not moving, not mixing and therefore less likely to mix with water that has not lost enough calories or is still slightly above zero (Celsius), thereby redistributing heat to prevent freezing. It also usually freezes at the surface where much colder air will help suck the calories out of the surface layer. It is easier for the water to freeze when not moving. It can grow on other solids, especially the exposed rocks or dirt of the bank which have already lost more heat and are below freezing; and then on the ice shelf as it builds up.

Another mechanism would be as the water sloshes onto the ice shelf, it is exposed to the below-freezing air and can create an additional thin layer of ice by losing heat to the air, so it builds bit by bit.

In still water, there is an inversion that happens. Due to the peculiarity of water molecule shapes, water expands when it get below 4C, so almost-ice is lighter (less dense) than warmer water to a point. the entire pond will cool to 4C with hot water rising, after which the cold layer becomes the top layer, there’s no more mixing, and the top layer gets colder and eventually loses enough heat to the air and freezes. Meanwhile, the lower levels of the pond stay about 4C.

I work on the Ohio River. It takes extremely low flow and a long cold snap to make it freeze. Any current brings slightly warmer water to the surface due to turbulence. When I was a kid it froze solid enough that some people walked across it (about 1000 feet).

We occasionally get floods with freezing temperatures. Backwater areas with no flow will freeze over. When the water recedes there are sections of bottomland woods with sheets of ice 8 feet up in the trees. Don’t walk underneath it because it can kill you when it collapses.

Assuming your river is like most and the deepest part is in the middle, the middle will flow the fastest and the edges the slowest. The slower moving water will freeze first because it is moving slowest and the area with the most movement, the middle will keep moving the longest until eventually freezing over.

The real magic happens when the floating pancake ice starts to snag along the edges. Then you get ice dams and things rapidly freeze up from there, at least, on the surface.

Yes thats it. If the water wasn’t really close to freezing, it would melt any existing ice. So the water in the river must have cooled to nearly freezing upstream…

Of course this (more often) occurs when the water is flowing slowly … or stopped.
When its been snowing for a while and the melting of it (warm ground ? ) stops.

Just an example I witnessed once, but upon a largish lake not far from Anaconda, Montana. Temps were well below 0F and my buddy asked me to go ice-fishing with him. Ice upon lake was at least three feet deep for the hundreds of yards we scouted to be sure of safety (had my ~10yr son tagging along,so were cautios to a high degree despite several dozen other fisher folk having their trucks and fishing ‘cabins’ hooked to vehicle. All were a hudred yards from shoreline and we were sure that us, being on foot with only a powered-drill to make hole, would be safe most anywhere.

A couple hours after we arrived, some guy showedd up and drove his truck w/ cabin in tow (similar to one in* Grumyp Old Men* movie) way out towards middle far more central to center of lake from all the other fishers, and promptly broke through ice and sank. Others, including us, were quick to get the driver out of water (chain of people laying on belly, etc) and provide warmth/recovery, etc. he ice in middle portion of lake obviously could not support weight of the less-than-other comparable truck/cabins that did not go anywhere near middle of lake. It had not been cold enough long enough for central area of lake to become thick enough to support the weight (outward freezing from shoreline, per se).

A very interesting/educational day, for sure. What was kinda funny is that as we were going back to shoreline to leave (after catching more trout than I could have ever thought possible in such conditions, fun!) my lightweight broke through ice with one leg as ice was thinning as closeness to shoreline approached. He found a spot that was only inches thick, and water was only knee-deep at that point, with unfrozen water of about a foot in width at that point of shoreline. It was very overcast/dim visibility so little sun-warmth upon shore area, so we were very surprised to find such thin ice where we had entered (that was solid to bottom visually when we first arrived). The ground at shoreline and a few feet inland was frozen solid like permafrost, too.

Weird. And it cost that fool who sank his new truck, lots of money to recover (by law) his vehicle/cabin, I bet. Point is, the closer to middle of frozen area, flowing or not, the more likely its less thick than towards shore, IME. I observed the same on many different lakes/rivers in Idaho/Wyoming, as well.