Tom MacWood,
When I first got into the business, I was told to use nearby railroad grades (especially the bottom of bridges crossing major creeks, which are said to be 2 feet above the calculated high flood level) as a guide to 100 year flood elevations. That data was probably available, as most track building in the US started iafter the civil war, and peaked in 1917. And, although more sophisicated now, the railroad engineers (the kind that draw plans, not the kind that run trains) had some data and used it conservatively.
Ross could have used that, or local knowledge from the old guys (Yeah, Mr. Ross, in the flood of 1888, the water got to the top of the first floor of the farmhouse, and that's the biggest flood I've seen in my 76 years on this planet) I still use that kind of field data, albeit with a bit more fear that some lawyer some day will argue that the Owner should have paid an engineer $100K to do a study to prove theoretically what actually happened in the real world!
However, as I have said before, Kris J. graciously allowed me to look at hundreds of Ross plans, and I never saw that mentioned, although I may not have been looking at floodplain courses. Many, many green plans in the Tufts archives call for getting fill to raise the back of the green from bunkers in front. It was the closest haul.
I also saw Ross filed notes for Franklin Hills in Detroit with Jeff Mingay, and on the par 3 third, he lowered the green for vision and also lowered another one in a side hill (Hole 2) to balance cut and fill, according to his field notes. All in all, pretty modern, and practical thinking. And evidence that, as Tom Doak says, Ross used all of a property.