Bart Schultz
The Enclosing Dam is 32.5 kilometres long and was for a long time the longest sea dike in the world until the 33.9 kilometres long sea dike of the Saemangeum Project was completed in South Korea in 2006. In the IJsselmeer, a water level of 0.20 m-NAP (metres below Amsterdam Ordnance Datum) is preferred from the beginning of April through September, and a level of 0.40 m-NAP from October through March. These levels were and still are fairly standard, but nowadays they are more flexible when dry or wet periods are expected.
In 1973, I obtained my civil engineering degree from Delft University of Technology. At the time I graduated, the dike from Lelystad to Enkhuizen had not yet been built, and the area of the IJsselmeer and its openly connected peripheral lakes was approximately 2,000 square kilometres. My thesis topic was the IJsselmeer as a stochastic system.
At that time, Delft University of Technology had a huge computer on which you could enter your own programs on punch cards. This was usually done in the afternoon. You could then check the next day whether the calculation had worked or whether you needed to adjust the stack of punch cards. Unthinkable now, but very advanced at the time. You could perform calculations yourself that, until then, would have been very time-consuming or even impossible. This enormous computer had a capacity of 80 kilobytes, a fraction of what your average mobile phone would have today.
In 1968, Rijkswaterstaat published the report The Water Management of the Netherlands. This report included a determination of the probability of exceeding the preferred water level of the IJsselmeer in winter if the dike from Lelystad to Enkhuizen had been closed, which would have reduced the area of the IJsselmeer to 1,200 square kilometres. This probability was determined using a statistical method. It was considered a good idea for a student to investigate whether with a new method a similar result would be obtained. Based on this, I started working on my engineers thesis.
If you view the IJsselmeer as a system, varying amounts of precipitation, river discharge, and pumped water from the adjacent polders enter, while evaporation and discharge occur at several sluices. The large sluices in the Enclosing Dam were the most important, but there were also sluices to the North Sea Canal, and the Provinces of North Holland and Friesland. Such a system involves variables in all of the aforementioned components.
The aim of my study was to use these variables to calculate the probability of exceedance of the preferred water levels, how much water could be available for the west and north of the country during dry periods, and to investigate the potential impact of closing the dike from Lelystad to Enkhuizen during the winter on exceedance of the water level of the IJsselmeer. An additional factor to consider was the operation of the weirs in the Nederrijn and Lek rivers, which allows additional water to flow through River IJssel to the IJsselmeer when the discharge of River Rhine is low. That is where stochastics (the science of probability and statistics) came into the picture.
I received all relevant available data from Rijkswaterstaat. This was a significant amount, as accurate measurements had been taken since the Enclosing Dam was closed. I started analysing the situation based on half-yearly figures. That was relatively manageable and yielded results relatively quickly. Then I switched to monthly figures. That's where the complications arose. While the monthly figures were more or less sufficient for determining potential water shortages, for winter water level exceedances, daily figures ultimately became necessary.
Ultimately, I managed to perform all calculations based on monthly figures, which already required a great deal of programming and computer work. Fortunately, it was considered sufficient to award me the Degree of Civil Engineer. Things were clear by then, before the summer break. To achieve any improvement in the results of the statistical approach, the calculations really had to be performed based on daily figures. That would have been very time-consuming, and even the computer at Delft University of Technology was not really capable of that at the time.
If a wet spell occurs in winter, as was the case at the time of writing this article in December 2023, I still checked how the water level in the IJsselmeer had developed. This can now be easily done on the Rijkswaterstaat website at waterinfo.rws.nl. Although the statistical line was determined before 1968, it still seems to be fairly accurate, even under the influence of changes in the Rhine River Basin and climate change. Perhaps someone would like to approach this stochastically. Modern computers should be more than capable to do this. With the measurement series now spanning 50 more years, it would also be interesting to use a much simpler statistical approach to examine how the fluctuations in the water in the IJsselmeer have developed.
