Towards Satellite Altimetry-Based Hydrological Modelling in Ungauged Catchments - A Water-Level Based Calibration of Rainfall-Runoff Models
Abstract
Traditional streamflow prediction requires model calibration using gauge discharge. However, most catchments in the world are ungauged, making hydrological analysis and water management difficult in these regions. A new calibration method that does not rely on gauge discharge is explored to assist in modelling for ungauged catchments. Based on an approximate monotonic relationship between discharge and water level integrated with rainfall-runoff models, the effectiveness of using water level data directly in the model calibration is investigated. Our approach has important implications to the direct use of satellite altimeter derived water level in rainfall-runoff model calibration.
We integrated a 3-parameter power function into general rainfall-runoff models to convert the output discharge to the water level. The calibration is based on the estimated water level from modified rainfall-runoff models and the observed time series of water level. Two water level time series are used in this study: (1) the ground-measured data in daily time step (benchmark scheme); (2) the synthetically generated satellite altimetry data with a temporal resolution of 10 days. In addition, to constrain the results, a few regionalised discharge signatures, such as the mean and the 95th percentile of discharge, are included in the calibration as extra objective functions. The results show that by calibrating against the water level data directly, the models could reproduce the discharge dynamics (timing, peaks and fluctuations) with reasonable accuracy. And with the constraining effects of the regionalised discharge signatures, the benchmark scheme performs well with a median NSE value of 0.63 in 130 study catchments. When the synthetic altimetry data are used, the performances decreased slightly. Our calibration method performs better in the wet catchments, and this is associated with the superior performance of the discharge-based calibration and more accurate regionalised discharge constraint estimates over the corresponding catchments. Overall, to use water level data directly in rainfall-runoff calibration provides a promising future for the calibration in ungauged catchments, and the method will be extended to use satellite altimetry data in the future study.- Publication:
-
AGU Fall Meeting Abstracts
- Pub Date:
- December 2019
- Bibcode:
- 2019AGUFM.H44G..06J
- Keywords:
-
- 1816 Estimation and forecasting;
- HYDROLOGY;
- 1836 Hydrological cycles and budgets;
- HYDROLOGY;
- 1847 Modeling;
- HYDROLOGY;
- 1855 Remote sensing;
- HYDROLOGY