Does a large number of parameters enhance model performance? Comparative assessment of common catchment model structures on 429 catchments
Introduction
After more than 30 years of ‘classical’ hydrological model development, there is an increasing trend in hydrology today to use new tools, that provide distributed information of catchment characteristics. It is tempting to introduce newly available information into increasingly complex catchment models, when one is faced with the task of accurately representing the inherent complexity of real systems. However, although this approach might be useful in terms of knowledge of the processes, it has limitations when applied in an operational context. Conversely, simple catchment models that lump catchment heterogeneities and represent the transformation of precipitation into streamflow, conceptually or empirically, are generally easy-to-use tools with low data requirements. In spite of the crude approximation resulting from their lumped nature and simple structure, such models have proved efficient in many case studies reported in the literature and they are undoubtedly useful for engineers and water managers. Daily conceptual models (of interest here) have successfully fulfilled most operational requirements, such as flood frequency assessment (Cameron et al., 1999, Uhlenbrook et al., 1999), reservoir management (Yang et al., 1995), flood and drought forecasting (see, e.g. Yang and Michel, 2000). Although these models have, up to now, been unable to predict the change in streamflow caused by land-use changes, they are apt at objectively detecting such changes (see, e.g. Lavabre et al., 1993, Lørup et al., 1998).
Since the early 1960s, many hydrologists have concentrated their efforts on designing rainfall–runoff models. Because they have attracted such widespread interest and so many different ones are being developed, the need for comparative studies was expressed quite early on (WMO, 1975) to evaluate the ability of models to simulate streamflow and to provide guidelines for end-users. Model assessment however is a tricky exercise and the conclusions of such experiments generally depend on the methodology of the comparisons and the characteristics of the test catchments. This is all the more true as models are mostly tested on limited numbers of catchments. Due to increasing computational capacity, it is possible today to extensively test simple models against great many catchments, under a wide range of climate conditions. However it would be naı̈ve to believe that, one might, from a broad-based comparison, identify one single outstanding model according to any assessment criterion that would satisfy all water stakeholders (Leviandier, 1988). Nevertheless, we believe that the robustness and reliability of a model resides primarily in its ability to perform under as varied a set of hydrological conditions as possible. From this standpoint, comparative assessments serve to highlight strengths and weaknesses of modelling approaches of various complexity.
Section snippets
Existing comparative assessments
Comparative assessment of the performances of rainfall–runoff models is not a new issue but is rarely the focus of much research, whereas there is a plethora of studies reporting satisfactory results from a single model. Reviews of most comparison exercises carried out so far can be found in Michaud and Sorooshian (1994) or in Refsgaard and Knudsen (1996). The latter state that, in general, no firm conclusion can be drawn regarding differences in model performances. The conclusions of
Objective
The main objective of this study is to test the performances of several structures derived from well-known rainfall–runoff models on a large sample of catchments within a common framework. Our study was conducted in four steps:
- 1.
Collection of a large sample of data from a wide variety of catchments under different climate conditions.
- 2.
Selection of a variety of existing continuous lumped conceptual or empirical rainfall–runoff models working at a daily time-step. Simple versions with a limited
Catchments and data
Except for the studies by Vandewiele et al., 1992, Xu and Vandewiele, 1995, who assessed the performances of monthly water-balance models on, respectively, 79 and 91 basins in Belgium, China and Burma, previous comparisons generally applied the models to a small number of catchments (less than 10). Although computing limitations may have prevented extensive testing in the past, it is today easier to implement testing schemes that can accommodate a large amount of calculations. Here, models are
Model selection
The choice of models used in this comparative assessment was based on a wide review of the literature. Because this study is oriented towards operational hydrology, all tested models have low data requirements and can be readily used in an operational context. Excluded from the comparison are models that are either spatially distributed, event-based or ‘physically based’ (in the sense that they require field measurements). The study concentrates on lumped, continuous, empirical or conceptual
Testing methodology
As mentioned above, the aim of this study was not to assess original models and their attached modelling methodologies such as those proposed in ready-to-use hydrological packages. These methodologies are not only based on a rainfall–runoff model structure, but also include specific procedures, chosen by the modeller, to select calibration periods, determine the optimum values of parameters, assess model fit quality or to determine uncertainties. Here, our sole objective was to examine crude
Results and discussion
The 19 daily rainfall–runoff model structures, as well as the baseline M50 model, were successively applied to the 429 catchments of the sample. All 1284 calibration runs and their corresponding 3204 verification tests were performed for each structure. In the discussion of the results, we concentrate mainly on performances obtained in verification mode, since this is the most common mode of model use in an operational context. In the following, the term ‘model’ sometimes used alone stands for
Conclusion and future work
This paper discusses the degree of model complexity (as reflected by the number of optimised parameters) required to simulate rainfall–runoff relationships on a wide variety of catchments. To assess the actual value of complexity in a model, an extensive testing scheme was carried out on 19 daily lumped model structures with three to nine optimised parameters. Model structures were tested for their ability to simulate streamflow, i.e. the target variable of such models in an operational
Acknowledgements
ENGEES (National engineering school for water and environment in Strasbourg, France) is thanked for its support of this study. The authors thank Dr Francis H.S. Chiew at the Department of Civil and Environmental Engineering of the University of Melbourne, Australia, for providing data sets of the Australian catchments and for his fruitful review of this paper; Dr Eric Servat at the Research Institute for Development (IRD, formerly ORSTOM) in Montpellier, France, for providing data sets for
References (83)
- et al.
Estimation of the ARNO model baseflow parameters using daily streamflow data
Journal of Hydrology
(1999) Prophecy, reality and uncertainty in distributed hydrological modelling
Advances in Water Resources
(1993)- et al.
Flood frequency estimation by continuous simulation for a gauged upland catchment (with uncertainty)
Journal of Hydrology
(1999) - et al.
Application of the daily rainfall–runoff model MODHYDROLOG to 28 Australian catchments
Journal of Hydrology
(1994) - et al.
Comparison of six rainfall–runoff modelling approaches
Journal of Hydrology
(1993) - et al.
A procedure for the selection of objective functions for hydrologic simulation models
Journal of Hydrology
(1977) - et al.
Comparative analysis of several conceptual rainfall–runoff models
Journal of Hydrology
(1991) - et al.
Physical interpretation and sensitivity analysis of the TOPMODEL
Journal of Hydrology
(1996) - et al.
A criterion of efficiency for rainfall–runoff models
Journal of Hydrology
(1978) - et al.
Measurement and utilization of on-site soil moisture data
Journal of Hydrology
(1996)
Computation of the instantaneous unit hydrograph and identifiable component flows with application to two small upland catchments
Journal of Hydrology
Monte Carlo assessment of parameter uncertainty in conceptual catchment models: the Metropolis algorithm
Journal of Hydrology
Use of spatially distributed water table observations to constrain uncertainty in a rainfall–runoff model
Advances in Water Resources
Changes in the hydrological response of a small Mediterranean basin a year after a wildfire
Journal of Hydrology
Assessing the effects of land use change on catchment runoff by combined use of statistical tests and hydrological modelling: case studies from Zimbabwe
Journal of Hydrology
Operational estimates of actual evapotranspiration and their significance to the science and practice of hydrology
Journal of Hydrology
Uncertainty in rainfall–runoff model simulations and the implications for predicting the hydrologic effects of land-use change
Journal of Hydrology
River flow forecasting through conceptual models. Part I — a discussion of principles
Journal of Hydrology
Methods for combining the outputs of different rainfall–runoff models
Journal of Hydrology
Application of an empirical infiltration equation in the SMAR conceptual model
Journal of Hydrology
The ARNO rainfall–runoff model
Journal of Hydrology
Multiple nonlinear statistical models for runoff simulation and prediction
Journal of Hydrology
Methodology and comparative study of monthly models in Belgium, China and Burma
Journal of Hydrology
The interdependence and applicability of some statistical quality measures for hydrological models
Journal of Hydrology
Parsimonious monthly rainfall–runoff models for humid basins with different input requirements
Advances in Water Resources
The ASCE task committee on definition of criteria for evaluation of watershed models of the watershed management committee, Irrigation and Drainage division, Criteria for evaluation of watershed models
Journal of Irrigation and Drainage Engineering
Principles and confidence in hydrological modelling
Nordic Hydrology
The future of distributed models: model calibration and uncertainty prediction
Hydrological Processes
A physically based, variable contributing area model of basin hydrology
Hydrological Sciences Bulletin
Lumped catchment models
Mathematical models of catchment behavior
American Society of Civil Engineers Proceedings
Hierarchical testing of three rainfall–runoff models in small forested catchments
Journal of Hydrology
Effective and efficient global optimization for conceptual rainfall–runoff models
Water Resources Research
GR3J: a daily watershed model with three free parameters
Hydrological Sciences Journal
Automatic calibration of conceptual rainfall–runoff models: optimization algorithms, catchment conditions, and model structure
Water Resources Research
Effects of model complexity and structure, data quality and objective function on hydrologic modeling
Journal of Hydrology
Cited by (479)
Hydrological models weighting for hydrological projections: The impacts on future peak flows
2023, Journal of Hydrology