Search

The paper presents prominent Nordic contributions to stochastic methods in hydrology and water resources during the previous 50 years. The development in methods from analysis of stationary and independent hydrological events to include non-stationarity, risk analysis, big data, operational research and climate change impacts is hereby demonstrated. The paper is divided into four main sections covering flood frequency and drought analyses, assessment of rainfall extremes, stochastic approaches to water resources management and approaches to climate change and adaptation efforts. It is intended as a review paper referring to a rich selection of internationally published papers authored by Nordic hydrologists or hydrologists from abroad working in a Nordic country or in cooperation with Nordic hydrologists. Emerging trends in needs and methodologies are highlighted in the conclusions. [ABSTRACT FROM AUTHOR]

Traditional continuous hydrological models have a large number of free parameters whose values need to be determined through calibration, and thus their applicability is limited to gauged basins. For prediction in ungauged catchments, hydrologists generally follow regionalization methods to develop region-specific calibration-free continuous models. An alternative attempt was made recently to develop a calibration-free model by proposing an empirically derived universal 'decay function' that enables definition of instantaneous dryness index as a function of antecedent rainfall and solar energy. The model was earlier tested in the USA, and its performance was found to be comparable to that shown by regionalization-based models. Here, we test the instantaneous dryness index-based calibration-free model considering data from 108 Indian catchments. The medians of coefficient of determination (R2), Nash-Sutcliffe efficiency (NSE) and Kling-Gupta efficiency (KGE) values for the study catchments, respectively, are 0.50, 0.38 and 0.40. Furthermore, the model's performance significantly improved upon Box-Cox transformation (RBC 2, NSEBC and KGEBC, respectively, are 0.70, 0.52 and 0.57), suggesting that the model predicts discharge quite well except during flood periods. Overall, our results suggest the model can be used as an alternative platform for predicting discharge in ungauged catchments in the USA and peninsular India, if not in every part of the world. [ABSTRACT FROM AUTHOR]

Water plays a central part in many aspects of life, and good hydrological science is needed to inform many decisions that affect human health and livelihoods. Hydrological scientists are aware of the importance of their work, and make great efforts to develop relevant research and advice. Decisionmakers want to use the best science in their decisions, but rarely read academic papers or attend conferences. This gap is often filled either by organisations that aim to bridge the science to policy gap (boundary organisations) or hydrological consultants. Academic hydrologists can aid the efficient transfer of knowledge to practitioners in these organisations by writing review papers, participating in professional conferences and engaging practitioners in their work. Researchers should also try to anticipate future evidence requirements and prepare research that will meet these. Faced with hydrological problems, decision-makers should continue to seek the right professional advice, and hydrological scientists should respect the expertise of decision-makers in making decisions. [ABSTRACT FROM AUTHOR]

Hydrologists are often faced with the problem of missing values in a precipitation-runoff process database to construct runoff prediction models. They tend to use simple and naive methods to deal with the problem of missing data. Thus far, the common practice has been to discard observations with missing values. In this paper, we present some statistically principled methods for gap filling and discuss the pros and cons of these methods. We employ and discuss imputations of missing values by means of self-organizing map (SOM), multilayer perceptron (MLP), multivariate nearest-neighbor (MNN), regularized expectation-maximization algorithm (REGEM) and multiple imputation (Ml) in the context of a precipitation-runoff process database in northern Iran in order to construct a serially complete database for analyses such as runoff prediction. In our case, the SOM and MNN tend to give similar and robust results. REGEM and Ml build on the assumption of multivariate normal data, which we don't seem to have in one of our cases. MLP tends to produce inferior results because it fragments the data into 68 different models. Therefore, we conclude that it makes most sense to use either the computationally simple MNN method or the more demanding SOM. [ABSTRACT FROM AUTHOR]

In urban water design and management, many hydrologic problems involve the frequency characterisation of rainfalls. Hydrologists are commonly asked to evaluate rainfall intensities for given recurrence frequencies or to indicate how rare an observed event is by estimating its return period. This study aims to improve the characterisation of rainfall hazard over a city by linking point to areal rainfall frequency analysis. We use a stochastic rainfall field generator based on the turning-bands method directly to assess areal rainfall distributions and to illustrate the link between different approaches. The simulating algorithm is applied to rainfall data from the city of Marseilles. The frequency analysis of simulated fields provides the elements to deal with the notions of return period and severity of observed storm events. The study concludes on the importance of a unified approach to assess rainfall better. [ABSTRACT FROM AUTHOR]