Uniqueness of place and process representations in hydrological modelling

By Keith Beven. 2000

565 citations recorded by the source corpus.

Abstract. This paper addresses the problem of uniqueness of catchment areas in relation to model representations of flow processes. The uniqueness of field measurements as a limitation on model representations is discussed. The treatment of uniqueness as a residual from a modelled relationship may conceal information about the uniqueness of catchments, while the treatment of uniqueness as a set of parameter values within a particular model structure is problematic due to the equifinality of model structures and parameter sets. The analysis suggests that a fully reductionist approach to describe the uniqueness of individual catchment areas by the aggregation of descriptions of small scale behaviour will be impossible given current measurement technologies. A suggested strategy for the representation of uniqueness of place as a fuzzy mapping of the landscape into a model space is suggested. This will lead to a quantification of the uncertainty in predictions of any particular location in a way that allows a conditioning of the mapping on the basis of the available data. This process can incorporate a hypothesis testing approach to model evaluation but the problem of multiple behavioural models may provide an ultimate limitation on the realism of process representations: not on the principle of realism but on the possibility of unambiguous process representations.

Subfields: Environmental Engineering, Ocean Engineering, Water Science and Technology

View Paper: https://www.semanticscholar.org/paper/72b001be1e042c3bd954ce1bf730946e0b1e7920

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Environmental Engineering

In Ocean Engineering

In Water Science and Technology

Citations in the corpus

Nonparametric direct mapping of rainfall‐runoff relationships: An alternative approach to data analysis and modeling? (2004, methodology): “The recourse to input and output data for parameter estimation appears as unavoidable for presently available models [Beven, 2000, 2002].”
Physics‐based hydrologic‐response simulation: Seeing through the fog of equifinality (2006, background): “…the equifinality problem for hydrology and related disciplines (also see Haines-Young and Petch, 1983; Culling, 1987; Aronica et al., 1998; Hankin and Beven, 1998; Schulz et al., 1999; Beven, 2000, 2002a,b; Brazier et al., 2000; Beven and Freer, 2001; Hreiche et al., 2002; Binley and Beven, 2003).”
Dalton Medal Lecture: How far can we go in distributed hydrological modelling? (2001, background): “However, my own view is that there is actually an opportunity here to put hydrological prediction on a firmer scientific basis (see Beven, 2000a).”
Ensemble flood forecasting: a review. (2009, background): “Spatial stationarity cannot be assumed as each catchment is unique (Beven, 2000).”
Modelling everything everywhere: a new approach to decision-making for water management under uncertainty (2012, background): “Epistemic uncertainties are an inherent part of modelling environmental systems and result in what Beven (2006) called non-ideal cases.”
A manifesto for the equifinality thesis (2006, background): “Ambiguity has also been used to reflect model identification problems in a variety of ways .”
Searching for the Holy Grail of scientific hydrology: Q t =( S, R, Δt ) A as closure (2006, background): “…is possible at all) would be the deﬁning act of scientiﬁc hydrology (it would not be the end of scientiﬁc hydrology as there would still be some interesting problems associated with applying that solution to di ﬀ erent applications in di ﬀ erent places, see, for example, Beven, 2000, 2002).”
A comparison of similarity indices for catchment classification using a cross-regional dataset (2012, background): “Uniqueness of place [3] tends to limit our ability to create generalizable hypotheses about the overall functioning of hydrological systems, an issue which has been described as ‘‘one of the most vexing’’ in hydrology [4, p.”
Pursuing the method of multiple working hypotheses for hydrological modeling (2011, methodology): “[11] Although hydrological models are often recognized as hypotheses of catchment behavior [e.g., Beven, 2001; Kuczera and Franks, 2002; Andréassian et al., 2009], a model is more akin to an assemblage of coupled hypotheses.”
Subsurface flow contribution in the hydrological cycle: lessons learned and challenges ahead—a review (2013, unknown): “broad consensus that such heavily parameterized models lead to high predictive uncertainty (Beven 2000). One more issue that is addressed well by Brunner et al. (2012) is the worth of observation data in identification of parameters (parameter identifiability) and predictive uncertainty.”
What Role Does Hydrological Science Play in the Age of Machine Learning? (2020, background): “As summarized by Hrachowitz et al. (2013), “ Beven (2000) highlighted the varying importance of different hydrological processes, active at different time scales in different catchments, and thereby emphasized uniqueness of place as a consequence of the variability of nature.”
Flow Prediction in Ungauged Catchments Using Probabilistic Random Forests Regionalization and New Statistical Adequacy Tests (2019, background): “, 2011; Westerberg & Birkel, 2015), and the quality of regionalization procedures (Beven, 2000; Wagener & Montanari, 2011).”
Towards an alternative blueprint for a physically based digitally simulated hydrologic response modelling system (2002, background): “This means that computational improvements are not only a matter of resolving finer characteristics of the flow but also of characterising of the medium or channels in which the flow is taking place (Beven, 2000).”
Moving beyond heterogeneity and process complexity: A new vision for watershed hydrology (2007, background): “Consequently, the resulting models are heavily overparameterized, and many combinations of parameter values can yield the same final result, leading to a large degree of predictive uncertainty [Beven, 2000].”