On typical range, sensitivity, and normalization of Mean Squared Error and Nash-Sutcliffe Efficiency type metrics

Hoshin Vijai Gupta, Harald Kling

Research output: Contribution to journalArticle

41 Citations (Scopus)

Abstract

We show that Mean Squared Error (MSE) and Nash-Sutcliffe Efficiency (NSE) type metrics typically vary on bounded ranges under optimization and that negative values of NSE imply severe mass balance errors in the data. Further, by constraining simulated mean and variability to match those of the observations (diagnostic approach), the sensitivity of both metrics is improved, and NSE becomes linearly related to the cross-correlation coefficient. Our results have important implications for analysis of the information content of data and hence about inferences regarding achievable parameter precision.

Original languageEnglish (US)
Article numberW10601
JournalWater Resources Research
Volume47
Issue number10
DOIs
StatePublished - 2011

Fingerprint

mass balance
normalisation
parameter
analysis

ASJC Scopus subject areas

  • Water Science and Technology

Cite this

On typical range, sensitivity, and normalization of Mean Squared Error and Nash-Sutcliffe Efficiency type metrics. / Gupta, Hoshin Vijai; Kling, Harald.

In: Water Resources Research, Vol. 47, No. 10, W10601, 2011.

Research output: Contribution to journalArticle

@article{a6231bd83f7a435cac712c239cdb10b8,
title = "On typical range, sensitivity, and normalization of Mean Squared Error and Nash-Sutcliffe Efficiency type metrics",
abstract = "We show that Mean Squared Error (MSE) and Nash-Sutcliffe Efficiency (NSE) type metrics typically vary on bounded ranges under optimization and that negative values of NSE imply severe mass balance errors in the data. Further, by constraining simulated mean and variability to match those of the observations (diagnostic approach), the sensitivity of both metrics is improved, and NSE becomes linearly related to the cross-correlation coefficient. Our results have important implications for analysis of the information content of data and hence about inferences regarding achievable parameter precision.",
author = "Gupta, {Hoshin Vijai} and Harald Kling",
year = "2011",
doi = "10.1029/2011WR010962",
language = "English (US)",
volume = "47",
journal = "Water Resources Research",
issn = "0043-1397",
publisher = "American Geophysical Union",
number = "10",

}

TY - JOUR

T1 - On typical range, sensitivity, and normalization of Mean Squared Error and Nash-Sutcliffe Efficiency type metrics

AU - Gupta, Hoshin Vijai

AU - Kling, Harald

PY - 2011

Y1 - 2011

N2 - We show that Mean Squared Error (MSE) and Nash-Sutcliffe Efficiency (NSE) type metrics typically vary on bounded ranges under optimization and that negative values of NSE imply severe mass balance errors in the data. Further, by constraining simulated mean and variability to match those of the observations (diagnostic approach), the sensitivity of both metrics is improved, and NSE becomes linearly related to the cross-correlation coefficient. Our results have important implications for analysis of the information content of data and hence about inferences regarding achievable parameter precision.

AB - We show that Mean Squared Error (MSE) and Nash-Sutcliffe Efficiency (NSE) type metrics typically vary on bounded ranges under optimization and that negative values of NSE imply severe mass balance errors in the data. Further, by constraining simulated mean and variability to match those of the observations (diagnostic approach), the sensitivity of both metrics is improved, and NSE becomes linearly related to the cross-correlation coefficient. Our results have important implications for analysis of the information content of data and hence about inferences regarding achievable parameter precision.

UR - http://www.scopus.com/inward/record.url?scp=80055000807&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80055000807&partnerID=8YFLogxK

U2 - 10.1029/2011WR010962

DO - 10.1029/2011WR010962

M3 - Article

AN - SCOPUS:80055000807

VL - 47

JO - Water Resources Research

JF - Water Resources Research

SN - 0043-1397

IS - 10

M1 - W10601

ER -