Large-sample assessment of varying spatial resolution on the streamflow estimates of the wflow_sbm hydrological model

Abstract

Abstract. Distributed hydrological modelling moves into the realm of hyper-resolution modelling. This results in a plethora of scaling-related challenges that remain unsolved. To the user, in light of model result interpretation, finer-resolution output might imply an increase in understanding of the complex interplay of heterogeneity within the hydrological system. Here we investigate spatial scaling in the form of varying spatial resolution by evaluating the streamflow estimates of the distributed wflow_sbm hydrological model based on 454 basins from the large-sample CAMELS data set. Model instances are derived at three spatial resolutions, namely 3 km, 1 km, and 200 m. The results show that a finer spatial resolution does not necessarily lead to better streamflow estimates at the basin outlet. Statistical testing of the objective function distributions (Kling–Gupta efficiency (KGE) score) of the three model instances resulted in only a statistical difference between the 3 km and 200 m streamflow estimates. However, an assessment of sampling uncertainty shows high uncertainties surrounding the KGE score throughout the domain. This makes the conclusion based on the statistical testing inconclusive. The results do indicate strong locality in the differences between model instances expressed by differences in KGE scores of on average 0.22 with values larger than 0.5. The results of this study open up research paths that can investigate the changes in flux and state partitioning due to spatial scaling. This will help to further understand the challenges that need to be resolved for hyper-resolution hydrological modelling.

Hyper-resolution PCR-GLOBWB: opportunities and challenges from refining model spatial resolution to 1 km over the European continent

“…While the 50k_50k run yields negative median KGE values, this picture improves for finer resolutions, with the 1k_1k and 1k_10k runs comfortably exceeding the accuracy threshold of −0.41, as defined by Knoben et al (2019). Higher skill with finer resolution is in line with previous research (Altenau et al, 2017), although this should not be expected to be the case for all stations due to locality effects in the scaling process (Aerts et al, 2022). As in our case, only around 7 % of all stations do not show an improvement at all, but roughly 70 % of the stations receive their highest KGE values for the 1k_1k run (Fig.…”

“…O'Neill et al (2021) applied and evaluated the hydrological model ParFlow (Maxwell et al, 2015) configured over the CONUS at 1 km spatial resolution, covering the years 2003 until 2006, constituting first-of-its-kind studies covering such an extensive area at true hyper-resolution. Aerts et al (2022) assessed changes in model accuracy of the hydrological model wflow_sbm (Schellekens et al, 2020) at multiple resolutions below 1 km 2 for the CAMELS dataset (Catchment Attributes and MEteorology for Large-sample Studies; Newman et al, 2015;Addor et al, 2017), again solely covering the CONUS. While the spatial resolution applied is hyper-resolution, the lack of either spatially continuous simulations over large extents or short simulation periods makes it challenging to define a substantial benchmark that is representative to the realm of GHMs.…”

“…Except for the precipitation forcing, all simulations make use of the same input and parameter sets and are downscaled or upscaled only if necessary. In our analysis, we focus not merely on simulated discharge, as done by Aerts et al (2022), for instance, but also on terrestrial water storage anomalies, evaporation, and soil moisture to provide a more holistic picture of model performance than what is typically the norm (see O'Neill et al, 2021). Such a very first thorough analysis of a hyper-resolution GHM is in the first place not intended to claim that hyper-resolution should be the new normal for GHMs but rather to gather opportunities and challenges of the scale dependencies of refining the spatial resolution of the model and its forcing.…”

Modelling and forecasting daily streamflow with reservoir operationin the upper Chao Phraya River basin, Thailand

“…which demonstrates its potential to improve the realism of the hydrological process and parameter representation. It has also shown good performance in a broad range of applications (e.g., López López et al, 2016;Hassaballah et al, 2017;Giardino et al, 2018;López López, 2018;Gebremicael et al, 2019;Rusli et al, 2021;Aerts et al, 2022). Therefore, it was selected for this thesis.…”

Hyperlocal Modeling: Reality or Wishful Thinking?

“…Estos modelos tienen como objetivo lograr una representación de los procesos hidrológicos en esa área particular. Dicho objetivo involucra un análisis que busca una mayor comprensión de los fenómenos hidrológicos a nivel local, haciendo posible datos de salida con mayor resolución, lo cual puede implicar un aumento en la comprensión de la compleja interacción de heterogeneidad dentro del sistema hidrológico (Aerts et al, 2022).…”