Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling

Gupta, H. V.; Kling, Harald; Yılmaz, Koray K.; Martínez, Guillermo

doi:10.1016/j.jhydrol.2009.08.003

Cited by 4,007 publications

(3,098 citation statements)

References 38 publications

Supporting

Mentioning

2,691

Contrasting

Unclassified

Order By: Relevance

“…The effect of initial conditions can be longer in drier climates. The parameter sets were considered behavioural as soon as they fulfilled a criterion that minimizes the Euclidean distance between observations and simulations for three components: the correlation, the relative variability, and the relative bias, the KlingGupta efficiency (E KG , Gupta et al, 2009). …”

Section: Hydrologiska Byråns Vattenbalansavdelning Model (Hbv)mentioning

confidence: 99%

Mapping (dis)agreement in hydrologic projections

Melsen

Addor

Mizukami

et al. 2018

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

Abstract. Hydrologic projections are of vital socioeconomic importance. However, they are also prone to uncertainty. In order to establish a meaningful range of storylines to support water managers in decision making, we need to reveal the relevant sources of uncertainty. Here, we systematically and extensively investigate uncertainty in hydrologic projections for 605 basins throughout the contiguous US. We show that in the majority of the basins, the sign of change in average annual runoff and discharge timing for the period 2070-2100 compared to 1985-2008 differs among combinations of climate models, hydrologic models, and parameters. Mapping the results revealed that different sources of uncertainty dominate in different regions. Hydrologic model induced uncertainty in the sign of change in mean runoff was related to snow processes and aridity, whereas uncertainty in both mean runoff and discharge timing induced by the climate models was related to disagreement among the models regarding the change in precipitation. Overall, disagreement on the sign of change was more widespread for the mean runoff than for the discharge timing. The results demonstrate the need to define a wide range of quantitative hydrologic storylines, including parameter, hydrologic model, and climate model forcing uncertainty, to support water resource planning.

show abstract

Section: Hydrologiska Byråns Vattenbalansavdelning Model (Hbv)mentioning

confidence: 99%

Mapping (dis)agreement in hydrologic projections

Melsen

Addor

Mizukami

et al. 2018

Hydrol. Earth Syst. Sci.

View full text Add to dashboard Cite

show abstract

“…Based on that, and analysing the simulation period, the NSE and PBIAS suggest that both models have a "good" performance, except the UD model, where the NSE presents a "satisfactory" performance assessment. iii) Reviewing the DGA and UD models during the calibration period, it is observed NSE values are worse during the simulation period, suggesting that the NSE is affected by its sensitive to the longitude of the observed data series and to the magnitude of bias (see McCuen et al [27], Gupta et al [23]). It is important to point out that the parameter A (parameter for adjustment the precipitation data) was defined during the calibration as 1.13 and 1.58 for DGA and UD rainfall inputs respectively.…”

Section: Resultsmentioning

confidence: 99%

“…The analysis and discussion were based on the simulation results, and the following model performance assessment functions: i) the NSE, ii) the percent of bias (PBIAS) (described on Gupta et al [22]), iii) the Kling-Gupta efficiency (Gupta et al [23]), and iv) the relative root mean squared error (RRMSE).…”

Section: Calibration Validation and Simulationmentioning

confidence: 99%

The use of global gridded datasets in a hydrological model for a scarce-data Andean watershed

Muñoz

Billib²,

Arumí

et al. 2011

WIT Transactions on Ecology and the Environment

View full text Add to dashboard Cite

Global gridded climatological (GGC) datasets including precipitation, temperature, pressure, and circulation indices, among others are becoming more and more precise, accessible and common in climate and hydrological research. In this study, we evaluate these datasets as an alternative input to supply the lack of measured climatological data in a Chilean Andean watershed in order to develop a monthly water balance model. A conceptual model was carried out for the simulation of stream-flows in the mountainous area of the Polcura River basin in south-central Chile. Based on 18 years of simulation and four model performance assessments, we concluded that from GGC datasets it is possible to reproduce observed flows and the snow-rain regime with "good" performance, being an adequate alternative to supply the lack of measured data, but taking the following considerations i) the GGC rainfall datasets in the Andean area are undervalued; this effect is transferred to the simulated flows, and must be fixed through an amplification coefficient (a calibration parameter), ii) due to the nature of the GGC datasets and to the orographic effect produced by the Andean mountains, the rainfall amounts are damped during the rainy season which result in a sub-estimation of the simulated flows during the winter, and limits the model scope to applications where peak flows are dispensable, and iii) the simulated snow-melt minimum and mean flows are close to the observed flows, which suggest that the GGC datasets are an adequate alternative for estimating the evapotranspiration and snow-melting amounts.

show abstract

“…It is a comprehensive stochastic algorithm which contains elements of calibration and uncertainty analysis. The Kling-Gupta efficiency (KGE) [GUPTA et al 2009] was used as an objective function. The KGE can range from -∞ to 1, where 1 is optimal.…”

Section: Multi-site Calibrationmentioning

confidence: 99%

Challenges in modelling of water quantity and quality in two contrasting meso-scale catchments in Poland

Marcinkowski

Piniewski

Kardel

et al. 2016

Journal of Water and Land Development

View full text Add to dashboard Cite

For citation: Marcinkowski P., Piniewski M., Kardel I., Srinivasan R., Okruszko T. 2016. Challenges in modelling of water quantity and quality in two contrasting meso-scale catchments in Poland. Journal of Water and Land Development. No. 31 p. 97-111. DOI: 10.1515/jwld-2016. AbstractThis study presents an application of the SWAT model (Soil and Water Assessment Tool) in two meso--scale catchments in Poland (Upper Narew and Barycz), contrasting in terms of human pressures on water quantity and quality. The main objective was multi-variable and multi-site calibration and validation of the model against daily discharge, sediment and nutrient loads as well as discussion of challenges encountered in calibration phase. Multi-site calibration and validation gave varied results ranging from very good (daily discharge) to acceptable (sediment, nitrogen and phosphorus loads in most of gauges) and rather poor (individual gauges for all variables) in both catchments. The calibrated models enabled spatial quantification of water yield, sediment and nutrient loads, indicating areas of special concern in terms of pollution, as well as estimation of contribution of pollution from different sources, indicating agriculture as the most important source in both catchments. During the calibration process a number of significant issues were encountered: (i) global vs. local parametrization, (ii) simulation of different pools of water quality parameters in reservoirs and streams and (iii) underestimation of NO 3 -N loads in winter due to farmers practices. Discussion of these issues is hoped to aid SWAT model users in Poland in a deeper understanding of mechanisms of multi-variable and multi-site calibration.

show abstract

Decomposition of the mean squared error and NSE performance criteria: Implications for improving hydrological modelling

Abstract: 22The mean squared error (MSE) and the related normalization, the Nash-Sutcliffe efficiency (NSE

Cited by 4,007 publications

References 38 publications

Mapping (dis)agreement in hydrologic projections

Mapping (dis)agreement in hydrologic projections

The use of global gridded datasets in a hydrological model for a scarce-data Andean watershed

Challenges in modelling of water quantity and quality in two contrasting meso-scale catchments in Poland

Contact Info

Product

Resources

About