Skip to main content
ARS Home » Plains Area » Temple, Texas » Grassland Soil and Water Research Laboratory » Research » Publications at this Location » Publication #319795
Title: Applications of explicitly-incorporated/post-processing measurement uncertainty in watershed modeling

Author
item YEN, HAW - Texas A&M Agrilife
item HOQUE, YAMEN - Texas A&M Agrilife
item WANG, XIUYING - Texas A&M Agrilife
item Harmel, Daren

Submitted to: Journal of the American Water Resources Association
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 12/22/2015
Publication Date: 4/30/2016
Citation: Yen, H., Hoque, Y., Wang, X., Harmel, R.D. 2016. Applications of explicitly-incorporated/post-processing measurement uncertainty in watershed modeling. Journal of the American Water Resources Association. 52(2):523-540.

Interpretive Summary: The importance of measurement uncertainty in terms of evaluating model predictions has been recently stated in the literature. The impact of measurement uncertainty indicates the potential vague zone in the field of watershed modeling where the assumption that measured data have no uncertainty may not be appropriate. It is important to note here that the measurement uncertainty refers to the uncertainty in measured data such as flow and nutrient concentrations that are used to evaluate model output. Although measurement uncertainty can affect model calibration results, it is rarely incorporated in modeling practice. Measurement uncertainty can be incorporated in two schemes: one during model calibration and two after calibration. In this study, both schemes are implemented in a case study of the Arroyo Colorado Watershed, Texas, USA, to investigate potential differences in model predictions. Unexpectedly, no substantial differences were observed for flow predictions. Measurement uncertainty did not cause dramatic differences in most sediment and ammonia predictions. However, results were affected in cases with measurement uncertainty greater than 50%. Therefore, it is concluded that measurement uncertainty may not have a major impact on model predictions until certain threshold is reached.

Technical Abstract: The importance of measurement uncertainty in terms of calculation of model evaluation error statistics has been recently stated in the literature. The impact of measurement uncertainty on calibration results indicates the potential vague zone in the field of watershed modeling where the assumption that measured data are deterministic may not be appropriate. It is important to note here that the measurement uncertainty refers to the uncertainty in measured data such as flow and nutrient concentrations that are used to evaluate model output. Although measurement uncertainty can affect model calibration results, it is rarely incorporated in modeling practice. Measurement uncertainty can be incorporated in two schemes: explicitly-incorporated (MU-EI) during model calibration and post-processed (MU-PP) after calibration is completed. In this study, both schemes are implemented in a case study of the Arroyo Colorado Watershed, Texas, USA, to investigate potential differences in model predictions. Unexpectedly, no substantial differences were observed for flow predictions. Measurement uncertainty did not cause dramatic differences in most sediment and ammonia predictions. However, calibration results were affected in cases with measurement uncertainty greater than 50%. Therefore, it is concluded that measurement uncertainty may not have a major impact on model predictions until certain threshold is reached. This study demonstrates that high levels of uncertainty in measured calibration/validation data can significantly affect parameter estimation, especially in the auto-calibration process.