The development of a new RTC system for a polder landscapes whose geography and current management practices developed over a time span of hundreds of years is a complex undertaking. The results to be achieved are generally formulated in terms of policies instead of clear requirements. Both in the literature on controller design and in practice the full process needed to transform policy statements into a control system that best serves those policies is somewhat neglected. Part of the problem is the rapid development of technology in general and resulting new options for automatic control. This is especially problematical in an environment where this path should be traversed together with the stakeholders. An analysis is presented of present Dutch practice and a proposal is made for a new approach that may well be applicable to RTC development for other complex water systems.

A polder-boezem system consists of a (large) number of polders that pump their drainage water into a network of watercourses and lakes. A few large pump stations then pump the drainage water into rivers or the sea. In some cases a sluice gate is used when water levels allow it. A receding horizon control algorithm for a polder-boezem system is presented where the selection of the control action is based on a prescribed order of use for the different states of available pumps and sluices. It is an extension of an earlier algorithm that operated boezem pump stations only. A further extension is proposed that will allow inclusion of the control of polder pumps. Arguments are presented to support the claim that, from the point of view of transparency, this algorithm is better suited to the control of polder-boezem systems than receding horizon control with a traditional objective function.

A first prototype of a Python package that can generate input for, perform runs of, and the analyse output of the Delft3D FM 1D2D simulation software corresponding to both simple and complex canal networks is demonstrated. The package can, for instance, generate input files for a canal system with a wide range of friction values and a range of variations on the design cross sections and structures used. It can then start parallel Delft3D runs for these input files. Finally, results can be analysed from Python. Results are presented for the rise time associated with a step change in the discharge for a prismatic canal with a trapezoidal cross section and either a weir or a gate at the downstream end. The size of the step change, the parameters of the cross section, the friction, and the canal slope are varied.

The dependency structure between hydrological variables is of critical importance to hydrological modelling and forecasting. When a copula capturing that dependence is fitted to a sample, information on the uncertainty of the fit is needed for subsequent hydrological calculations and reasoning. A new method is proposed to report inferential uncertainty in a copula parameter. The method is based on confidence curves constructed with the use of a pseudo maximum likelihood estimator for the copula parameter. The method was tested on synthetic data and then used as a tool in two hydrological examples. The first examines the probability of major floods in two locations on the Rhine River and its tributaries in the same calendar year. In the second example, rainfall–runoff from a karst region in Tunisia was analysed to determine a confidence interval for the delay between precipitation and runoff.

The representation of uncertainty in results is an important aspect of statistical techniques in hydrology and climatology. Hypothesis tests and point estimates are not well suited for this purpose. Other statistical tools, such as confidence curves, are better suited to represent uncertainty. Therefore three parametric methods to construct confidence curves for the location of a sudden change in the properties of a time series, a change point (CP), are analyzed for three distributions: log-normal, gamma, and Gumbel. Two types of change are considered: a change in the mean and a change in the standard deviation. A question that confidence curves do not answer is how likely the null hypothesis of ‘no change’ is. A possible statistic to help answer this question, denoted by Un, is introduced and analyzed. It is compared to the statistic that underlies the Pettitt test. All methods perform well in terms of coverage and confidence set size. One method is based on the profile likelihood for a CP, the other two, first defined in this article, on the pseudolikelihood for a CP. The main advantage of the pseudolikelihood over the profile likelihood lies in the much lower computational cost. The confidence curves generated by the three methods are very similar. In a limited test on time series of measurements found in the literature, the methods gave results that largely matched those reported elsewhere. Some results are also given for an order one autoregressive series with a lognormal marginal distribution.

Networks of open channels form an important category of environmental systems. They are used not only to transport irrigation and drainage water, but also as highways for barges transporting raw materials and goods. Automatic control of these systems poses specific problems. A local stability analysis for an open canal that is split into several parts by sluice gates under discrete time control is proposed. Theoretical justification is provided, and the method is tested for a simple controller. The method allows the examination of local stability of a series of canals when equipped with a controller from a large class, linear and non-linear. The analysis is based on the analysis of the eigenvalues of a matrix derived from the controlled system that is small enough to allow for parameter optimization.

In almost all practical applications of control, technological and economical consid -erations impose limits on communication speed, frequency of communication, and frequency of actuator adjustment. Such limits turned the analysis of sampled data systems into a flourishing field. Water systems pose a particular challenge: the systems are networks of canals and reservoirs spread over large areas, and the actuators are relatively large and exposed to the elements. In this study, a theorem on the local exponential stability of sampled data systems with variable control time step and variable delay in the communication between the non-linear continuous time process and the non-linear discrete time controller is presented. To illustrate the application of the theorem, it is applied to a simple water system.

Management of water systems is becoming more and more complex; this creates opportunities for the application of control theory. These opportunities are the subject of a course on operational water management given to students of the water management department, Delft University of Technology, over the past 15 years. Traditional examples in control theory courses are taken from industry and do not easily map to water systems, so examples were developed that use water systems to illustrate control theory concepts. This provided the students with a link between control theory and water management practice.

The problem of smoothing dry weather inflow variations for a Waste Water Treatment Plant (WWTP) that receives sewage from multiple mixed sewer systems is formulated. A first rough control algorithm that uses branch and bound is presented. The control algorithm uses a form of Model Predictive Control. Trials showed that the algorithm had trouble satisfying two constraints that were initially regarded as ‘soft’ constraints. As a result, a closer look was taken at the feasibility of the problem. A family of simpler problems were derived to do so. These auxiliary problems made it possible to show that for a given subclass feasibility was strongly dependent on the choice of problem parameters.

Several commonly-used nonparametric change-point detection methods are analysed in terms of power, ability and accuracy of the estimated change-point location. The analysis is performed with synthetic data for different sample sizes, two types of change and different magnitudes of change. The methods studied are the Pettitt method, a method based on the Cramér von Mises (CvM) two-sample test statistic and a variant of the CUSUM method. The methods differ considerably in behaviour. For all methods the spread of estimated change-point location increases significantly for points near one of the ends of the sample. Series of annual maximum runoff for four stations on the Yangtze River in China are used to examine the performance of the methods on real data. It was found that the CvM-based test gave the best results, but all three methods suffer from bias and low detection rates for change points near the ends of the series.

Arguments are presented in favor of modeling sewer systems and in particular Dutch sewer systems as a sampled data system with events. Basic limitations on controlling these systems when ignoring their hybrid nature are stated. The traditional control scheme for the Dutch systems is given as an example of event driven local control. Basic limitations on systems using a sampled data approach without an event driven component are derived. To provide context a brief description of a sampled data controller for a sewer system based on set-point tracking is given. This is followed by an explanation of how the absence of event driven control limits its effectiveness.