Research papersAssessing future runoff changes with different potential evapotranspiration inputs based on multi-model ensemble of CMIP5 projections
Introduction
Runoff is one of the key processes in water transport both for surface water bodies (e.g., rivers, lakes, wetlands, and oceans) and groundwater (Ghasemizade and Schirmer, 2013, Kuchment, 2004). The amount of runoff from each rainfall event has direct or indirect influences on water availability in many aspects of human activities such as agricultural and industrial production, and domestic life (Allan et al., 2020, Devi et al., 2015). Previous studies have shown that climate change with increased temperature and changed rainfall patterns has great impacts on runoff (Arnell and Gosling, 2013, Bosshard et al., 2013, Im et al., 2009). For instance, Arnell and Gosling (2013) found that more than 47% of the land surface would experience increases in mean annual runoff whereas around 36% of that would witness mean annual runoff decrease due to changes in temperature and rainfall. Meanwhile, considerable variation in the impact of climate change on runoff has been found among different regions (Arnell and Gosling, 2013, Do et al., 2017, Shen et al., 2014). Providing a robust projection of regional runoff under climate change plays a significant role in understanding local water resources management and revealing the impacts of a changing climate on local hydrological cycle (Allan et al., 2020, Yan et al., 2020).
Different kinds of methods such as climate elasticity (Xing et al., 2018, Yang and Yang, 2011), Bayesian approach (Freni et al., 2009), and hydrological model (Chiew et al., 2018, Islam et al., 2014) can be used for runoff simulation and projection. Compared with other methods, one of the most important advantages of hydrological models is that they are capable to detect the hydroclimate response to changes by offering a comprehensive and reliable approach at a certain catchment (Guo et al., 2017a, Li et al., 2020). In previous studies, hydrological models were widely used as powerful tools to investigate runoff response to climate change (Fowler et al., 2018, Pechlivanidis et al., 2016, Vaze and Teng, 2011). Specifically, historical observed rainfall and runoff sequences are used to calibrate and validate the performance of hydrological models. Then, runoff can be projected by the calibrated hydrological models forced with the climatic factors derived from global climate models (GCMs) (Arnell, 2011, Chen and Yu, 2015). For instance, Senent-Aparicio et al. (2017) investigated the influence of climate change on runoff in Mediterranean Europe with SWAT model and found that runoff would decrease due to the increase in temperature and decrease in rainfall. In Western Australia, Islam et al. (2014) downscaled climatic data from 11 GCMs under A2 and B1 emission scenarios to drive LUCICAT model for the future rainfall-runoff projection. They found that projected decrease in rainfall would result in a large decrease in runoff for Western Australia in the mid and late of 21st century.
According to closed water balance, runoff for a certain region is roughly the difference between rainfall and actual evapotranspiration (ETa) in a long-term period (Montaldo and Oren, 2018). Therefore, the estimation of ETa is expected to influence the simulation of runoff (Riegger and Tourian, 2014). In the process of runoff simulation with most hydrological models, potential evapotranspiration (ETp) is an essential input to calculate ETa used for simulating runoff (Bai et al., 2016, Li and Zhang, 2017). However, various ETp models generally produce different ETp estimates (Feng et al., 2016, Kumar et al., 1987; Kumar Roy et al., 2020). In this case, which ETp model could produce better runoff simulation is an important question to answer (Dakhlaoui et al., 2020, Oudin et al., 2005, Seiller and Anctil, 2016). In other words, will the difference in ETp estimates result in different runoff simulations/projections? Addressing this question is important in runoff projection under a changing climate as ETp is greatly influenced by future climate (Pan et al., 2015, Zheng et al., 2017).
Oudin et al. (2005) adopted ETp estimated by 27 ETp models to drive four rainfall-runoff hydrological models and investigate the influence of different ETp inputs on historical runoff simulation over 308 catchments across France, Australia, and the United States. They found that these hydrological models showed low sensitivity to ETp inputs but temperature-based and radiation-based ETp models yielded the best runoff simulation. Under future climate scenarios, Seiller and Anctil (2016) assessed the sensitivity of 20 hydrological models in runoff projection to ETp estimated by 24 different equations. They found that the different ETp inputs exerted moderate influence on runoff projection. In Korea, Bae et al. (2011) investigated the sensitivity of three hydrological models to seven ETp methods in runoff projection with downscaled climate data from 13 GCMs. They concluded that the influence of different ETp on runoff projection became larger. On the contrary, Dakhlaoui et al. (2020) found that discharge simulated by three hydrological models was not sensitive to the ETp estimates under different climate conditions. In summary, though studies about the influence of different ETp inputs on future runoff projection are becoming common, there is no consistent conclusion yet.
Another unavoidable challenge in future runoff projection is the uncertainty caused by many factors such as GCMs, hydrological models (Knutti and Sedláček, 2012, Teng et al., 2015), and emission scenarios (Woldemeskel et al., 2016). For instance, Teng et al. (2012) projected runoff based on 15 GCMs with five hydrological models in southeast Australia and found that uncertainty caused by GCMs was much larger than that caused by hydrological models. Vetter et al. (2017) investigated the uncertainty in runoff projection caused by five GCMs, four RCPs, and nine hydrological models across 12 large-scale catchments worldwide. They found that GCMs and RCPs were the main factors resulting in the uncertainty. Similarly, Chegwidden et al. (2019) found that the choice of RCPs or GCMs was the main source influencing the spread in annual streamflow volume and timing. These studies would provide useful information to quantify the dominant source of uncertainty in runoff projection. However, few of them considered the possible contribution of different ETp inputs and their interaction with other factors to the uncertainty in runoff projection. Given that ETp is essential for runoff projection, especially the influence of ETp may become larger under future climate scenarios (Seiller and Anctil, 2016), it is necessary to include ETp in the uncertainty analysis in runoff projection.
Thus, the objectives of this study are dual: 1) to investigate the influence of different ETp inputs both in historical runoff simulation and future runoff projection; 2) to quantify the relative contribution of GCMs, RCPs, ETp models, and their interaction to the uncertainty in runoff projection. To achieve these goals, we calibrated and validated Xinanjiang (XAJ), a rainfall-runoff hydrological model driven by different ETp inputs against observed historical runoff at a humid catchment in northeastern Australia. Then, we used validated XAJ model to project future runoff under RCP4.5 and RCP8.5 with climate data downscaled from 34 GCMs. Based on the projected runoff, we quantified the relative contribution of different factors with the method of analysis of variance. We expect this study can offer further insights into the impacts of different ETp inputs on runoff projection and help to clarify the role of ETp inputs on the related uncertainty. The knowledge from this study will be helpful to guide the ETp model choice in future runoff projection. Results in this study can also provide a way forward to narrow down uncertainty in runoff projection.
Section snippets
Study area
The study area is North Johnstone catchment (17°16′ S – 17°38′ S, 145°28′ E – 146°40′ E, Fig. 1), locating in the Wet Tropics of Queensland, Australia. It covers an area of 924 km2, with elevation ranging from 18 m to 1370 m (Zhang et al., 2020). The mean maximum and minimum temperatures in this catchment are around 26.0 °C and 16.7 °C, respectively. Mean annual rainfall in this catchment is around 2530 mm and mean annual runoff is around 1900 mm. The area is influenced by the monsoon and
Calibration and validation of the XAJ model
As one of the key inputs in runoff simulation, difference was observed in ETp estimated by different models (Figure S1). For instance, more ETp estimated by Ab was lower than 4 mm day−1 whereas JH estimated ETp were more likely to be higher than 4 mm day−1. However, the observed runoff and simulated runoff from the XAJ model (driven by different ETp inputs) did not show great difference, as indicated by similar R2 (Fig. 4, top panel), NSE (Fig. 4, middle panel), and RMSE (Fig. 4, bottom panel)
Low sensitivity of XAJ model to different ETp inputs
Compared with XAJ model driven by physically-based Penman calculated ETp, it showed comparable (or even better) ability in runoff simulation with temperature-based (HS) and radiation-based ETp inputs. For instance, with the same R2 and NSE, RMSE produced by XAJ model with Penman-ETp was even larger (0.05 mm day−1) than that produced by XAJ model with Ab and HS ETp inputs (Fig. 4). In spite of the small difference among different ETp inputs, XAJ model performed well in reproducing daily observed
Conclusions
We investigated the influence of different ETp inputs on runoff simulation and projection with XAJ model in North Johnstone catchment, northeast Australia. Meanwhile, we quantified the contribution of GCMs, RCPs, ETp models, and their interaction to the uncertainty in runoff projection with a method of three-way analysis of variance. Our findings indicated that XAJ model performed well in runoff simulation, in the study catchment. We found that XAJ model with different ETp inputs projected
CRediT authorship contribution statement
Lijie Shi: Conceptualization, Data curation, Formal analysis, Methodology, Visualization, Writing – original draft, Writing – review & editing. Puyu Feng: Data curation, Investigation, Formal analysis, Software, Writing – original draft. Bin Wang: Conceptualization, Methodology, Supervision. De Li Liu: Data curation, Methodology, Resources, Supervision. Hong Zhang: Data curation, Software. Jiandong Liu: Writing – review & editing. Qiang Yu: Funding acquisition, Investigation, Resources, Project
Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.
Acknowledgements
The first author acknowledges a scholarship from the Chinese Scholarship Council, and the NSW Department of Primary Industries provided office facilities for conducting this work. We acknowledge the support from Natural Science Foundation of China (No.41961124006, 41730645) and US National Science Foundation (1903722). We also appreciate the Scientific Information for Land Owners and the Bureau of Meteorology for the access of climatic data used in this study. Bernie Dominiak reviewed an early
References (85)
- et al.
Global performance ranking of temperature-based approaches for evapotranspiration estimation considering Köppen climate classes
J. Hydrol.
(2015) - et al.
The impacts of climate change on river flow regimes at the global scale
J. Hydrol.
(2013) - et al.
Hydrologic uncertainties in climate change from IPCC AR4 GCM simulations of the Chungju Basin, Korea
J. Hydrol.
(2011) - et al.
Uncertainties in runoff projections in southwestern Australian catchments using a global climate model with perturbed physics
J. Hydrol.
(2015) - et al.
Evaluation and comparison of CMIP6 and CMIP5 model performance in simulating the seasonal extreme precipitation in the Western North Pacific and East Asia
Weather Clim. Extremes
(2021) - et al.
A global-scale investigation of trends in annual maximum streamflow
J. Hydrol.
(2017) - et al.
Assessing the differences in sensitivities of runoff to changes in climatic conditions across a large basin
J. Hydrol.
(2011) - et al.
Trends in water quality in a subtropical Australian river-estuary system: responses to damming, climate variability and wastewater discharges
J Environ Manage
(2020) - et al.
Impacts of climate change on streamflow and floodplain inundation in a coastal subtropical catchment
Adv. Water Resour.
(2021) Combined effects of urbanization and climate change on watershed evapotranspiration at multiple spatial scales
J. Hydrol.
(2020)
Comparison of ELM, GANN, WNN and empirical models for estimating reference evapotranspiration in humid region of Southwest China
J. Hydrol.
Urban runoff modelling uncertainty: Comparison among Bayesian and pseudo-Bayesian methods
Environ. Modell. Software
Trend and magnitude of changes in climate variables and reference evapotranspiration over 116-yr period in the Platte River Basin, central Nebraska–USA
J. Hydrol.
Using spatial interpolation to construct a comprehensive archive of Australian climate data
Environ. Modell. Software
Uncertainty assessment of climate change impacts for hydrologically distinct river basins
J. Hydrol.
Is ET often oversimplified in hydrologic models? Using long records to elucidate unaccounted for controls on ET
J. Hydrol.
Uncertainty assessment of multi-parameter, multi-GCM, and multi-RCP simulations for streamflow and non-floodplain wetland (NFW) water storage
J. Hydrol.
Impacts of projected climate change on runoff in upper reach of Heihe River basin using climate elasticity method and GCMs
Sci. Total Environ.
Separating effects of vegetation change and climate variability using hydrological modelling and sensitivity-based approaches
J. Hydrol.
Predicting runoff in ungauged catchments by using Xinanjiang model with MODIS leaf area index
J. Hydrol.
Regionalising rainfall-runoff modelling for predicting daily runoff: Comparing gridded spatial proximity and gridded integrated similarity approaches against their lumped counterparts
J. Hydrol.
Which potential evapotranspiration input for a lumped rainfall–runoff model?
J. Hydrol.
An assessment of the severity of recent reductions in rainfall and runoff in the Murray-Darling Basin
J. Hydrol.
Using ensembles of adaptive neuro-fuzzy inference system and optimization algorithms to predict reference evapotranspiration in subtropical climatic zones
J. Hydrol.
Projecting potential evapotranspiration change and quantifying its uncertainty under future climate scenarios: A case study in southeastern Australia
J. Hydrol.
Attributing uncertainty in streamflow simulations due to variable inputs via the Quantile Flow Deviation metric
Adv. Water Resour.
Projection of future runoff change using climate elasticity method derived from Budyko framework in major basins across China
Global Planet. Change
Impacts of future climate change on water resource availability of eastern Australia: A case study of the Manning River basin
J. Hydrol.
Using an improved SWAT model to simulate hydrological responses to land use change: a case study of a catchment in tropical Australia
J. Hydrol.
Assessing the ability of potential evapotranspiration models in capturing dynamics of evaporative demand across various biomes and climatic regimes with ChinaFLUX measurements
J. Hydrol.
Advances in understanding large-scale responses of the water cycle to climate change
Ann. N. Y. Acad. Sci.
Uncertainty in the relationship between climate forcing and hydrological response in UK catchments
Hydrol. Earth Syst. Sci.
Quantifying the sources of uncertainty in an ensemble of hydrological climate-impact projections
Theor. Appl. Climatol.
Assessment of the influences of different potential evapotranspiration inputs on the performance of monthly hydrological models under different climatic conditions
J. Hydrometeorol.
Quantifying uncertainty sources in an ensemble of hydrological climate-impact projections
Water Resour. Res.
Impact of downscaled rainfall biases on projected runoff changes
Hydrol. Earth Syst. Sci.
How Do Modeling Decisions Affect the Spread Among Hydrologic Climate Change Projections? Exploring a Large Ensemble of Simulations Across a Diversity of Hydroclimates
Earth's Future
Impact assessment of climatic and land-use changes on flood runoff in southeast Queensland
Evaluation de l'impact des changements climatiques et de l'utilisation des terres sur l'écoulement de crue dans le Sud-Est du Queensland.
Rainfall-runoff modelling considerations to predict streamflow characteristics in ungauged catchments and under climate change
Water
Sensitivity of discharge projections to potential evapotranspiration estimation in Northern Tunisia
Reg. Environ. Change
A Review on Hydrological Models
Aquat. Procedia
Cited by (16)
Probabilistic assessment of drought impacts on wheat yield in south-eastern Australia
2023, Agricultural Water ManagementAssessment of the potential impacts of climate changes on Syr Darya watershed: A hybrid ensemble analysis method
2023, Journal of Hydrology: Regional Studies