Vernon
Now the programs were designed so that you probably could get access to intermediate
stages of the computation, but that's not really the way it was set up to run. You just put
it in there, and you got the answers out--that's what the planning people liked because they
didn't want to fool around with the intermediate stages. My argument, of course, was,
"Hey, how do you know something didn't go wrong in there." You may not even be able
to tell in your answer whether it's reasonably right or wrong if you haven't checked all
these different steps as you go along the way.
So the argument I use to get from Beard and Jay Frederick and some of the others that
were in the early end of that computer business was, "Hey, you can do all kinds of `what
So, if you really wonder if a
if' things with computers that you couldn't do before.
person can get better trained with a computer than they could by doing it by hand, they
can try out different things. If this doesn't work, try something else.
But I still had reservations about that. I felt that too many people, because they could stick
in data and get an answer, they didn't have to know anything about the theoretical
background of where all this came about. What is a unit hydrograph, what does it do?
How do you route floods through reservoirs and stuff like that? What are the procedures
in backwater studies? What are the formulas and so forth that go into the concepts that.
They get wrong answers, and they don't know it. All they know is that they got the
program to run, and they got an answer and they go ahead and apply it and also it kind of
avoided checking. They would just assume the computer is right. Where years ago when
they first started, when I first started, everything had to be checked. You never send
anything ahead without [checking]--if you had a computation sheet, it always had to be
computed by and checked by someone. You just never sent anything without checking.
Well, that kind of went by the wayside when the computers come along. You probably
still need that sort of thing though. I mean not necessarily checking every number in
there, but to be sure that whoever put the program together and the input data, knew what
they were doing.
Just to give you an example of one process that can get screwed up pretty easily and that's
flood frequency, flood flow frequency analysis. Even though they come up with Bulletin
which was put together by the interagency committee on hydrology, subcommittee
on hydrology, the procedure is fairly straightforward and it's not real simple but it's
straightforward and anybody can use it. But the problem is the data that goes into the
system.
Previous Page
