Revisiting good modelling practices and open workflows for decision support – where are we today and where to tomorrow?

Co-organized by EOS4

Convener: Diana SpielerECSECS | Co-conveners: Anneli GuthkeECSECS, Zhenyu WangECSECS, Catherine Moore, Dirk EilanderECSECS, Wouter Knoben

The approaches and methods we choose for a hydrological modelling study affect our modelling results and conclusions, and hence also their usefulness for decision support. As of today there is no common and consistently updated guidance on what good modelling practice is, and how we can achieve transparent, robust and reproducible workflows. While many useful practices such as scripted workflows, model benchmarking, controlled model comparisons, careful selection of calibration periods and methods, or testing the impact of subjective modelling decisions along the modelling chain exist, none of these can be considered common practice yet.

This session therefore intends to provide a platform for a visible and ongoing discussion on what ought to be the current standard(s) for an appropriate modelling protocol that considers uncertainty in all its facets and promotes transparency in the quest for robust and reliable results. We invite presentations of worked examples and software tools: What did(n’t) work? How were challenges overcome? How did developed workflows allow for detailed scrutiny of the techniques, assumptions, and interpretations of data, models, and their uncertainties? Contributions should aim to improve the scientific basis of (parts of) the modelling chain and put good modelling practice in focus again. This might include (but is not limited to) contributions on:

(1) Benchmarking to increase trust in model results
(2) Developing robust calibration and evaluation frameworks to improve transparency
(3) Going beyond common metrics in assessing model performance and realism
(4) Developing frameworks that enable hypothesis testing or consideration of alternative conceptual models
(5) Investigating subjectivity and documenting choices along the modelling chain
(6) Developing modelling protocols and/or scripted workflows to improve efficiency and reproducibility
(7) Examples of adopting the FAIR (Findable, Accessible, Interoperable and Reusable) principles in the modelling chain
(8) Methods for uncertainty analysis, data assimilation, and management optimization under uncertainty, e.g. in the decision-support context
(9) Communicating model results and their uncertainty to end users of model results
(10) Evaluating implications of model limitations and identifying priorities for future model development and data acquisition planning