The Applicability of SWOT’s Non-Uniform Space–Time Sampling in Hydrologic Model Calibration

Cassandra Nickles,Dongmei Feng,Edward Beighley

doi:10.3390/rs12193241

Cassandra Nickles, Dongmei Feng + Show 1 more

Open Access

https://doi.org/10.3390/rs12193241

Copy DOI

Abstract

The Surface Water and Ocean Topography (SWOT) satellite mission, expected to launch in 2022, will enable near global river discharge estimation from surface water extents and elevations. However, SWOT’s orbit specifications provide non-uniform space–time sampling. Previous studies have demonstrated that SWOT’s unique spatiotemporal sampling has a minimal impact on derived discharge frequency distributions, baseflow magnitudes, and annual discharge characteristics. In this study, we aim to extend the analysis of SWOT’s added value in the context of hydrologic model calibration. We calibrate a hydrologic model using previously derived synthetic SWOT discharges across 39 gauges in the Ohio River Basin. Three discharge timeseries are used for calibration: daily observations, SWOT temporally sampled, and SWOT temporally sampled including estimated uncertainty. Using 10,000 model iterations to explore predefined parameter ranges, each discharge timeseries results in similar optimal model parameters. We find that the annual mean and peak flow values at each gauge location from the optimal parameter sets derived from each discharge timeseries differ by less than 10% percent on average. Our findings suggest that hydrologic models calibrated using discharges derived from SWOT’s non-uniform space–time sampling are likely to achieve results similar to those based on calibrating with in situ daily observations.

Highlights

Hydrologic and hydrodynamic models are useful for assessing climate change impacts [1], predicting flood characteristics, forecasting applications [2], and understanding the transfer and storage of water and energy globally [3]
Kling-Gupta Efficiency (KGE) values increase on average by 0.01 when the model is calibrated using US Geological Survey (USGS) discharges sampled based on Surface Water and Ocean Topography (SWOT) overpasses as compared to the full daily record of discharges
Analyzing 10,000 model iterations shows that 93% of parameter iterations have KGE values with absolute percent differences less than 20% considering only SWOT temporal sampling

Summary

Introduction

Hydrologic and hydrodynamic models are useful for assessing climate change impacts [1], predicting flood characteristics, forecasting applications [2], and understanding the transfer and storage of water and energy globally [3]. Satellite remote sensing has been utilized to quantify and assess water security [4], and measurements can be coupled with hydrologic models to improve overall discharge estimation [5,6,7,8,9]. Calibration methods are used to optimize model parameters for estimating river discharge [10,11,12,13,14]. SWOT will observe river surface height, width, and slope for at least 90% of rivers wider than 50–100 m [15]. These measurements can be used in SWOT flow inversion algorithms to estimate discharge [16,17]. Ref. [19] shows that in the Mississippi River Basin, locations vary in being observed from zero to four times during the orbit cycle

Objectives

Methods

Results

Discussion

Conclusion