"Toronto experienced 100mm of rain yesterday"...
I'm having trouble properly conceptualizing this amount. I know that 100mm is about 4 inches, and I know that rain is measured in depth over a given area, so my rough intuition would suggest that there would be four inches of standing water on average across the terrain. This is clearly not what we're seeing in the photos, though, and I realize this conceptualization is also too simplistic - it fails to take into account pooling and runoff due to terrain. A mean terrain height would be under 4 inches of water, while lower areas would be under much more and higher areas would be under none. All in all, it's hard to picture large volumes of water like this in appropriate quantities spread out over appropriate areas. More effort is required to understand.
Wikipedia informs me that 1mm of rainfall is equal to 1 liter of water per square meter, so 100mm would be 100 liters per square meter. I don't know what 100 liters looks like, but I'm familiar with 2 liter bottles, so I can conceptualize 50 of those stacked up in a roughly 3 foot by 3 foot area. Unfortunately, the shape of the bottles leaves a lot of empty space inbetween them, so the whole heap takes up a lot more space than the liquid in the bottles would if poured out, and that screws up my mental picture. Let's try another frame of reference.
A quick google search shows me that a small, kitchen-sized trashcan holds about 50 liters, as does a large backpack. Working with the mental concept of two such trashcans or backpacks, I can easily imagine them sitting side by side in the 3 foot by 3 foot benchmark area. Most of the volume is vertical though, so I'll lay them down instead of standing them up. Still a lot of open space in those 9 square feet, though. At this point I have to start dividing the height of the containers and mentally filling the empty space with "slices" of the volume. I can picture it pretty well, though, and to my relief it seems to correspond to my initial calculations - the area would be entirely covered by about 4 inches of standing water.
So then how do we get the 4 feet of water we see in the linked picture? Presumably my earlier notion of pooling caused by the terrain explains it. I say presumably, because once again I can't quite conceptualize things properly, and have a hard time picturing how 4 inches of water spread out over a large area would end up as 4 feet of water in the path of that train.
Okay, so I've got two backpacks or kitchen trashcans of water in a 3 foot by 3 foot area. But then let's say I've got a combination of 10,000 such areas, or 90,000 square feet (roughly 10.000 meters square?). How much is that? Well, an American Football field (minus the endzones) is 48,000 square feet, so two of those next to each other. I can picture that pretty well.
Right. So. I've got 20,000 trashcans or backpacks of water. That's hard to conceptualize. Let's try a different tack. I've got 1,000,000 liters of water. How much is that? Well, an Olympic swimming pool holds 2,500,000 liters. You can fit 4 such pools in a single football field. Such pools are 2 meters in depth, or 6 feet. Working from that, I calculate... four inches of depth spread across the total area. Okay, yes, I knew that, but hey, this means I'm still on target.
But! Now I have a mass of water that I can conceptualize, and a region of land to match! Now I can mentally factor in terrain elevation. If we raise the edges of the area, like the curves of a bowl, such that half the total area is shifted upwards at least 4 inches, and do the opposite for the other half, lowering the center of the area at least 4 inches, and then add the water...
Okay, so that doubles the depth of the water in the lowered areas, and the raised areas remain high and dry. Repeat the process, subdividing the area into quarters instead of halves, and we double the depth at the lowest point again. So now we've got 16 inches of water pooled in an area 1/4 the size of the total region, in which there is only 16 inches of difference in elevation from the lowest to the highest points. And we've got a nice handy 1:1 ratio of height difference to water depth at the lowest point.
So in an 8 foot deep gulley, we'll see 4 feet of water. Finally, I can kind of picture this properly!