How Good Is Your Model Fit? Weighted Goodness‐of‐Fit Metrics for Irregular Time Series

Abstract

Weighted goodness-of-fit metrics may help to the evaluate model fit for head time series with irregular time steps.

Systematic visual analysis of groundwater hydrographs: potential benefits and challenges

“…The direct comparison approach by Haaf and Barthel (2018) is included. It is the authors' opinion that the advantage of visual classification is the ability to handle time series data that are inhomogeneous in terms of measurement intervals, start and end date, gaps and irregularities-all features typical of groundwater hydrographs (Collenteur 2021;Peterson et al 2017). Handling in this sense, not only means foremost detecting and characterizing irregularities, but also using data which otherwise is too poor for numerical evaluation.…”

Reply to Comment on "How Good is Your Model Fit? Weighted Goodness‐of‐Fit Metrics for Irregular Time Series"

“…In the model presented in Collenteur ( 2021a ) an AR(1) noise model was actually applied and irregular time steps were taken into account through its objective function (following von Asmuth and Bierkens 2005 ). However, as is clear from the example, we may still end up with a model that is biased toward the high (or even low) frequency period.…”

Comment on “How Good is Your Model Fit? Weighted Goodness‐of‐Fit Metrics for Irregular Time Series”

“…The weighting scheme of Collenteur provides a practical solution for this. However, the initial weights are not symmetrical in time: This can be improved by using instead: These initial weights still need to be normalized and made dimensionless by dividing by the sum of the initial weights (equation 4 in Collenteur 2021 ) before application.…”

“…As an illustration, I analyzed the time series of the same piezometer (from the Dutch national subsurface information database at https://www.DINOloket.nl/en/ ) as Collenteur ( 2021 ) with precipitation and Makkink evaporation series from the same meteorological stations of the Royal Dutch Meteorological Institute (KNMI) using the Metran software (Berendrecht and van Geer 2016 ; Zaadnoordijk et al 2019 ).…”

“…The technical commentary of Collenteur ( 2021 ) touches an important aspect in the use of groundwater head data in the conversion from manual measurements (and sampling) to sensors with automatic dataloggers (e.g., Post and von Asmuth 2013 ; Retike et al 2022 ). Collenteur offers a practical solution that improves evaluation for time series with a transition from regular manual measurements to high(er) frequency automatic logged groundwater heads.…”