Uncertainty in hydrological analysis of climate change: multi-parameter vs. multi-GCM ensemble predictions

Abstract

The quantification of uncertainty in the ensemble-based predictions of climate change and the corresponding hydrological impact is necessary for the development of robust climate adaptation plans. Although the equifinality of hydrological modeling has been discussed for a long time, its influence on the hydrological analysis of climate change has not been studied enough to provide a definite idea about the relative contributions of uncertainty contained in both multiple general circulation models (GCMs) and multi-parameter ensembles to hydrological projections. This study demonstrated that the impact of multi-GCM ensemble uncertainty on direct runoff projections for headwater watersheds could be an order of magnitude larger than that of multi-parameter ensemble uncertainty. The finding suggests that the selection of appropriate GCMs should be much more emphasized than that of a parameter set among behavioral ones. When projecting soil moisture and groundwater, on the other hand, the hydrological modeling equifinality was more influential than the multi-GCM ensemble uncertainty. Overall, the uncertainty of GCM projections was dominant for relatively rapid hydrological components while the uncertainty of hydrological model parameterization was more significant for slow components. In addition, uncertainty in hydrological projections was much more closely associated with uncertainty in the ensemble projections of precipitation than temperature, indicating a need to pay closer attention to precipitation data for improved modeling reliability. Uncertainty in hydrological component ensemble projections showed unique responses to uncertainty in the precipitation and temperature ensembles.

Reply to reviewer comment 1

“…If that was indeed true, then it would be justifiable to use any uncalibrated hydrological model to project future hydrological fluxes and state variables. However, a recent study demonstrated that while uncertainty of GCM projections was the dominant source for faster components of hydrologic response like surface runoff, the uncertainty of hydrological model parameterization was found to be a significant source of uncertainty, particularly for slow response components (Her et al, 2019). This finding indicates that calibration of hydrological models is still important…”

Simulating the Hydrological Processes of a Meso-Scale Watershed on the Loess Plateau, China

“…However, applying hydrological models in the real world is often challenged by calibration, validation, and predictability issues [7][8][9][10]. During the calibrated process, parameter sensitivity and uncertainty analysis often become a hurdle, even though hydrological models are normally calibrated to find optimal parameters set with the optimum objective functions [11][12][13]. Unless there is a careful sensitivity and uncertainty analysis, overestimation or underestimation of hydrologic regimes can cause over-design of mitigation measures or insufficient preparation for potential situation [13,14].…”

“…Although there have been many studies involving sensitivity analysis, uncertainty fitting and performance evaluation of SWAT model all over the world [2,11,14,16,18,35], identifying dominant parameters, reducing the magnitude of uncertainties for streamflow simulation and confirming actual hydrological processes still remain essential for some distinct regions. Uncertainties in hydrological modeling come from input data, model structure, and parameters [2,21].…”

“…Uncertainties in hydrological modeling come from input data, model structure, and parameters [2,21]. However, there are many key parameters in a watershed that describe the bio-geophysical characteristics and hydrological processes of the watershed, among which some parameters are hard to measure directly and can be gained only by empirical estimation and literature references, resulting in the increased uncertainties of the modeling system [11,17,21,36]. Several studies showed that using more informative data, modified model structures, or more parameter sets could reduce uncertainties [2,37,38].…”

The value of using hydrological datasets for water allocation decisions: earth observations, hydrological models, and seasonal forecasts

“…New hydrological datasets from earth observations, hydrological and land-surface models, as well as seasonal forecasts continue to be developed and tested around the world. The quality of these datasets has been improving rapidly due to the consideration of longer periods of record, increasingly finer spatial resolutions, better accuracy in hydrological estimates and predictions, and the consideration of uncertainty through ensembles (Derin et al, 2016;Dijk et al, 2016;Her et al, 2019;Orth et al, 2017;Schellekens et al, 2017). For example, a new version of MSWEP global precipitation dataset (historic dataset over 30 years) was enhanced from a 0.25 degree to a 0.1 degree spatial resolution in less than one year (Beck et al, 2017b(Beck et al, , 2018.…”