Validation of catchment models for predicting land-use and climate change impacts. 1. Method (original) (raw)
Related papers
Journal of Hydrology, 1996
Validation methods commonly used to test catchment models are not capable of demonstrating a model's fitness for making predictions for catchments where the catchment response is not known (including hypothetical catchments, and future conditions of existing catchments which are subject to land-use or climate change). This paper describes the first use of a new method of validation (Ewen and Parkin, 1996. J. Hydrol., 175: 583-594) designed to address these types of application; the method involves making 'blind' predictions of selected hydrological responses which are considered important for a particular application. SHETRAN (a physically based, distributed catchment modelling system) is tested on a small Mediterranean catchment. The test involves quantification of the uncertainty in four predicted features of the catchment response (continuous hydrograph, peak discharge rates, monthly runoff, and total runoff), and comparison of observations with the predicted ranges for these features. The results of this test are considered encouraging.
A framework for catchment prediction modelling
Models used in catchment prediction have often been developed for specific research problems or locations by individuals using software engineering practices that are now considered obsolete. The legacy of this is a range of models dealing with similar problems, using similar data input and output interpretation, but with a high diversity of operational features. Modern catchment management requires that policy development, planning and intervention be undertaken in an integrated fashion, with consideration given to physical, ecological, economic and social systems. Integrated catchment modelling offers managers some support in meeting these needs, although the range of different languages, platforms, and design approaches used in existing models means that integrated modelling can not be done by simply plugging together existing models. A modelling framework is being developed that covers not only module design and specification, but also relevant protocols for user involvement, testing and comparison, documentation, calibration and validation, and verification and peer review. This framework will be used by the Cooperative Research Centre for Catchment Hydrology as a basis for integration of appropriate existing modules and newly developed modules into a catchment prediction toolkit. A benefit of this approach is that researchers and modellers can spend more time on core module development, and less or no time on input and output management. Other benefits include a common software operational paradigm and consistent delivery approach for a suite of catchment management related software programs. It is anticipated that adoption of this framework will support the long term design and integration of a variety of modules relevant to the prediction of catchment behaviour.
Benefits and limitations of current approaches to whole of catchment modelling
Developments in catchment modelling frameworks, such as the new WaterCAST tool developed by the eWater CRC, have improved the ability to model catchment management and land use changes at the whole of catchment scale. While these tools have given modellers greater flexibility in the approaches, algorithms and frameworks which can be applied to a particular catchment, this paper will examine these improvements through a number of case studies and discuss current benefits and limitations when modelling with the WaterCAST tool at this scale.
Hydrological Sciences Journal, 2017
Assessing water resources is an important issue, especially in the context of climatic changes. Although numerous hydrological models exist, new approaches are still under investigation. In this context, we propose a modelling approach based on the physical principle of least action. We present new hypotheses to develop the model further, to widen its application. The improved version of the model MODHYPMA was applied on 20 sub-catchments in Africa and the USA. Its performance was compared with two well-known lumped conceptual models, GR4J and HBV. The model could be successfully calibrated and validated. In calibration, GR4J performed better, while other models had similar performance. In validation, MODHYPMA and GR4J performed similarly and better than HBV. The parameter λ has medium sensitivity while parameters λ and TX have low sensitivity. The parameter uncertainty for MODHYPMA, analysed using the GLUE methodology, was higher during high flows but with good p and r factors.
Spatially explicit versus lumped models in catchment hydrology – experiences from two case studies
NATO Science for Peace and Security Series C: Environmental Security, 2009
This paper analyses the major features of spatially explicit and lumped hydrological models based on two case studies. For two different catchments in West Africa and Germany model intercomparison studies were performed to reveal the model structure and spatial resolution dependent advantages and disadvantages of the different model types. It can be shown that different model types (lumped versus distributed models and conceptual versus physically based models) have benefits and drawbacks. But all model predictions of different type models contain some valuable information when used for the simulation of catchment water fluxes. Using local scale data from intense field experiments, the sophisticated and spatially explicit models simulate stream flow of a West African catchment with the same performance obtained by lumped models that can be calibrated more efficiently. In addition, the spatially explicit models generate plausible spatial patterns of state variables and processes which can be validated by additional observations. Using regional scale available data to predict stream flow of a German catchment, the simpler models tend to perform better in both calibration and validation periods. But while all models tend to show improved performance during the less extreme validation period, this improvement is greatest for some of the more complex models. Applying the same models (of different model types) to three land use change scenarios, there is broad agreement among the models on the expected hydrological change. This suggests that we can predict with some confidence the direction and magnitude of stream flow changes associated with 3 P.C. Baveye et al. (eds.), Uncertainties in Environmental Modelling land use change, especially by combining the predictions of different model types. As a short outlook, it is shown that a simple multi-model application offers a sound basis for multi-model ensembles that are based on a technique currently applied successfully in many atmospheric forecast and scenario studies.
Evaluation of catchment models
2003
Catchment models are by definition simplified representations of the real world system. This aggregation takes place in space and time and has several important consequences. First, there are no generally applicable rules to perform this aggregation, and the resulting model structure is usually a function of the modeller's hydrological understanding. Secondly, the model parameters cannot be measured directly in many cases, but have to be estimated.
Water Resources Research, 2012
This paper investigates the actual extrapolation capacity of three hydrological models in differing climate conditions. We propose a general testing framework, in which we perform series of split-sample tests, testing all possible combinations of calibration-validation periods using a 10 year sliding window. This methodology, which we have called the generalized split-sample test (GSST), provides insights into the model's transposability over time under various climatic conditions. The three conceptual rainfall-runoff models yielded similar results over a set of 216 catchments in southeast Australia. First, we assessed the model's efficiency in validation using a criterion combining the root-mean-square error and bias. A relation was found between this efficiency and the changes in mean rainfall (P) but not with changes in mean potential evapotranspiration (PE) or air temperature (T). Second, we focused on average runoff volumes and found that simulation biases are greatly affected by changes in P. Calibration over a wetter (drier) climate than the validation climate leads to an overestimation (underestimation) of the mean simulated runoff. We observed different magnitudes of these models deficiencies depending on the catchment considered. Results indicate that the transfer of model parameters in time may introduce a significant level of errors in simulations, meaning increased uncertainty in the various practical applications of these models (flow simulation, forecasting, design, reservoir management, climate change impact assessments, etc.). Testing model robustness with respect to this issue should help better quantify these uncertainties.
In this study, 17 hydrologists with different experience in hydrological modelling applied the same conceptual catchment model (HBV) to a Greek catchment, using identical data and model code. Calibration was performed manually. Subsequently, the modellers were asked for their experience, their calibration strategy, and whether they enjoyed the exercise. The exercise revealed that there is considerable modellers' uncertainty even among the experienced modellers. It seemed to be equally important whether the modellers followed a good calibration strategy, and whether they enjoyed modelling. The exercise confirmed previous studies about the benefit of model ensembles: Different combinations of the simulation results (median, mean) outperformed the individual model simulations, while filtering the simulations even improved the quality of the model ensembles. Modellers' experience, decisions, and attitude, therefore, have an impact on the hydrological model application and should be considered as part of hydrological modelling uncertainty.
Operational testing of a water balance model for predicting climate change impacts
Agricultural and Forest Meteorology, 1999
The ability of water balance models to incorporate month±month or seasonal variations in climate, snowfall and snowmelt, groundwater¯uctuations, soil moisture characteristics, and natural climatic variability makes them especially attractive for water resources studies of climatic changes. The use of conceptual models to explore the impact of climate changes has increased in recent years. Because of the success claimed for these studies, it is likely that computer simulation of catchments will increasingly be used by and for water resource managers as an aid to decision-making. There is therefore a need for a generally accepted method for demonstrating a model's ®tness for such use. The simple split-sample test method may be reasonable in the simplest case of the`®lling-in missing data' problem but certainly not if the express purpose of the model is to simulate records for conditions different from those corresponding to the calibration record, such as the problem with predicting the effects of climate changes where the data on the changed system are not (and cannot be) available for comparison with the model predictions. Thus, model validation must demonstrate`®tness for the said purpose'.