I thought of the news reports on a Friday or Saturday that predicted the Saint John River flood would peak the following Tuesday. And that happened as predicted. But I was surprised that the record-breaking nature of the flooding was not also predicted – a week, 10 days or two weeks beforehand. I can’t help but think that somebody knew or should have.
I’m certain there are well-developed and accurate models of flow in the St. John River and of contributing factors in the river’s watershed including snow cover and weather. These would be hydrologic and meteorologic models developed from empirical data collected in the watershed over many generations if not centuries. We have models in forensic engineering investigation that serve us well too.
Sounds technical but hydrology is simply the study of the flow of water in a watershed and meteorology is the study of weather. Put them together and you’ve got a powerful tool for predicting if a river is going to flood, when this will happen and how high.
A model is simply a set of ideas and numbers that describe the past, present or future of something such as an economy, business or, flow in a river. Models are built using measurements and observations – empirical data – of the things that characterize what you’re interested in.
A street map is a model. It shows the location of streets and other features of interest in an area. Things like businesses, buildings, the local coffee shop, etc. It doesn’t show things you’re not interested in like the height of the buildings and the level of the streets.
Good and accurate models:
- Fit the empirical data from which they’re built
- Explain past observations – like why the river flooded in the past
- Predict future observations – like when the river will flood again
- Are simple and inexpensive to use
Exhaustive data collection and study of flow in the River would have been done for design and construction of the Mactaquac Dam in 1968 – a few kilometres up river of Fredericton – and during the 50 years, half a century, after the Dam was operating. So why not an accurate prediction of the historic flooding – not an inch or so above all previous highs but something like a foot?
We rely on models in the forensic engineering investigation of the cause of foundation failure. Foundation design and construction is closely tied to the semi-empirical science of soil mechanics – a science partly based on measurements and observations and partly on theory. This science developed in the early 1900s and holds us in good stead in a forensic investigation.
Annual flooding in the St. John River was being recorded each year long before the early 1900s, and the weather – temperature and rainfall – and conditions in the watershed – snow pack – were being noted.
So, I would put my money on the existence of accurate models that would predict with a respectable degree of accuracy, a week or two in advance, that New Brunswick was going to be flooded-out. These models would include data on how the Mactaquac Dam is operated in storing and releasing water.
I was in contact with a professional forester who lives in Douglas a short distance up river of Fredericton. He wonders too about the flooding and also noted the Dam and it’s operation as a feature in the watershed.
I can’t help but wonder if a similar situation exists in British Columbia with the flooding there.
References
- Giere, Ronald N, Bickle,John, and Mauldin, Robert F., Understanding Scientific Reasoning, 5th ed., 2006, Thompson Nelson, Toronto
- Wikipedia, May 14 and 15, 2018