Can hydrological models benefit from using global soil moisture, evapotranspiration, and runoff products as calibration targets?

Yiwen Mei; Juliane Mai; Hong Xuan Do; Andrew Gronewold; Howard W. Reeves; Sandra M. Eberts; Richard G. Niswonger; R. Steve Regan; Randall J. Hunt

doi:10.1029/2022wr032064

USGS · 70241025

Can hydrological models benefit from using global soil moisture, evapotranspiration, and runoff products as calibration targets?

Yiwen Mei·Juliane Mai·Hong Xuan Do·Andrew Gronewold·Howard W. Reeves·Sandra M. Eberts·Richard G. Niswonger·R. Steve Regan·Randall J. Hunt

View original source ↗Export metadata ↓

Abstract

Hydrological models are usually calibrated to in-situ streamflow observations with reasonably long and uninterrupted records. This is challenging for poorly gage or ungaged basins where such information is not available. Even for gaged basins, the single-objective calibration to gaged streamflow cannot guarantee reliable forecasts because, as has been documented elsewhere, the inverse problem is mathematically ill-posed. Therefore, the inclusion of other observations, and the reproduction of other hydrological variables beyond streamflow, become critical components of accurate hydrological forecasting. In this study, six single- and multi-objective model calibration schemes based on different combinations of gaged streamflow, global-scale gridded soil moisture, actual evapotranspiration (ET), and runoff products are used for the calibration of a process-based hydrological model for 20 catchments located within the Lake Michigan watershed, of the Laurentian Great Lakes. Results show that the addition of gridded soil moisture to gaged streamflow in model calibration improves the ET simulation performance for most of the catchments, leading to the overall best-performing models. The monthly streamflow simulation performance for the experiments using gridded runoff products to inform the model is outperformed by those using the gaged streamflow, but the discrepancy is mitigated with increasing catchment scale. A new visualization method that effectively synthesizes model performance for the simulations of streamflow, soil moisture, and ET was also proposed. Based on the method, it is revealed that the streamflow simulation performance is relatively weak for baseflow-dominated catchments; overall, the 20 catchment models simulate streamflow and ET better than soil moisture.

Explore related subjects

Water Resources Research Article

Source-reported bounding extent: 41.43075831670021° to 46.27321789689552° latitude; -88.5330560089191° to -84.66751013004324° longitude. This indicates report coverage, not an exact sampling location. View area on OpenStreetMap.

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Yiwen Mei, Juliane Mai, Hong Xuan Do, Andrew Gronewold, Howard W. Reeves, Sandra M. Eberts, Richard G. Niswonger, R. Steve Regan, Randall J. Hunt. 2023-02-22. Can hydrological models benefit from using global soil moisture, evapotranspiration, and runoff products as calibration targets?. https://doi.org/10.1029/2022wr032064

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2026

A demonstrated workflow to efficiently identify better performing process-based model configurations

There are often competing options for process representation when designing process-based hydrologic models. Comparing candidate model configurations during model development can be computationally expensive and existing methods for evaluating prior-data conflict often cannot distinguish between multiple models with low prior-data conflict. Here, we demonstrate a workflow to efficiently identify better performing models by comparing prior model output distributions to observed data distributions using a metric referred to as Mahalanobis distance deviation (𝑀 D ). The 𝑀 D metric was developed to incorporate the notion of balancing model fit with model complexity into prior-data conflict approaches. Ninety-six different configurations of the Precipitation-Runoff Modeling System were designed for the East River watershed, Colorado by considering uncertainty in representing spatial discretization, forcing data, potential evapotranspiration, soil zone flow, snow depletion curve, and transpiration period processes. The prior-to-calibration adequacy of each model was estimated by calculating 𝑀 D for observed streamflow, areal runoff, actual evapotranspiration, groundwater recharge, soil moisture, and snow-covered area. All 96 models were then calibrated using the PEST++ iterative ensemble smoother to quantify their performance and validate the 𝑀 D results. The 𝑀 D metrics calculated from prior parameter ensembles as small as 25 exhibited strong correlation with calibrated objective function values highlighting the substantial information on model performance that can be gained prior to calibration through the workflow. Spatial discretization and forcing data were most influential on model performance and their better performing representations were most easily identified. This work demonstrates an efficient approach to identify better performing model configurations and the processes driving those improvements.

Water Resources Research↗

USGS2026

Evaluating SWOT satellite data with gage information in the United States and Canada

The Surface Water and Ocean Topography (SWOT) satellite mission represents a transformative approach to water resources management, offering high-resolution data on water-surface elevations (WSE) and extents. This study evaluates SWOT WSE accuracy in the United States and Canada with in situ streamgage and lake and reservoir gage data. We compare SWOT Level 2 Hydrology data products (Version D) to in situ timeseries of WSE on 335,531 paired observations across 2,629 gages. SWOT data accuracy and temporal coverage are explored as functions of various SWOT quality filters. After applying a standard set of defined filters and thresholds to compare paired measurements of SWOT and in situ observations (both referenced to the Earth Gravitational Model 2008), we find 1-σ (68th percentile) absolute differences of 20 and 11 cm, and median biases of 2.8 and 0.1 cm, for filtered river node and lake data, respectively. River reach and lake comparisons may be biased due to scale mismatches between in situ point-based data and spatially averaged SWOT data. Applying SWOT data quality filters improved median paired elevation R2 between unfiltered and filtered data from 0.12 to 0.92 across all reach, node, and lake observations. Overall, we find that 60.0% and 73.8% of all paired SWOT node and lake WSE observations that satisfy SWOT-observable criteria (ice-free observations from sufficiently large waterbodies within the SWOT cross track range) have elevations within 25 cm of reported gage observations. These comparisons also provide a baseline benchmark that can help inform continued SWOT algorithm refinement.

Water Resources Research↗

USGS2026

Big data for small streams: Leveraging a coarse national model to enhance fine-resolution stream temperature predictions

Predictive stream temperature models are commonly used to address spatial and temporal gaps in observational data because stream temperature information is essential for environmental management decisions. National models capture broad patterns driven by climate, topography, and seasonality, but often lack the spatial resolution to simulate fine-scale variation in groundwater discharge, riparian shading, or tributary inflows. In contrast, watershed models may simulate finer-resolution patterns but miss regional drivers or be limited by scarcity of observations. To generate watershed-extent, fine-resolution stream temperature predictions, we developed 18 multiscale machine-learning models that leverage (a) a broad-extent, coarse-resolution stream temperature model and (b) finer-resolution observations and predictor data. The coarse-resolution model simulates daily stream temperature across the conterminous United States on the National Hydrologic Geospatial Fabric (13.7 km median reach length; 865,364 km total). The fine-resolution models simulate daily stream temperature on the National Hydrography Dataset Plus version 2 (1.5 km median; 5,192,808 km total). We compared multiscale models to seven purely broad- (national) or watershed-extent models, testing them in eight watersheds (436–985 km 2 ) with differing hydrologic settings and under realistic data availability scenarios. Especially when observations within a watershed were scarce, multiscale models generated more accurate predictions than national or watershed-specific models. A recurrent graph convolution neural network pretrained on coarse predictions and fine-tuned on watershed-specific observations was most accurate across watersheds and data scenarios. These results highlight that leveraging both coarse, national results and watershed-specific data can increase model accuracies, particularly when observations within a watershed are spatially or temporally limited.

Water Resources Research↗