Publication Date:
2019
Description:
Abstract
One of the main problems of hydrologic/hydrodynamic routing models is defining the right set of parameters, especially on inaccessible and/or large basins. Remote Sensing techniques provide measurements of the basin topography, drainage system and channel width, however current methods for estimating riverbed elevation are not as accurate. This paper presents methods of altimetry data assimilation for estimating effective bathymetry of a hydrodynamic model. We tested past altimetry observations from satellites ENVISAT, ICESAT and JASON 2 and synthetic altimetry data from satellites ICESAT 2, JASON 3, SARAL and SWOT to assess future/present mission's potential. The data assimilation (DA) methods used were Direct Insertion, Linear Interpolation, the SCE‐UA optimization algorithm and an adapted Kalman Filter developed with hydraulically based variance and covariance introduced in this paper. The past satellite altimetry data assimilation was evaluated comparing simulated and observed water surface elevation (WSE) while the synthetic altimetry DA were assessed through a direct comparison with a “true” bathymetry. The SCE‐UA and hydraulically based Kalman Filter methods presented the best performances, reducing WSE error in 65% in past altimetry data experiment and reducing biased bathymetry error in 75% in the synthetic experiment, however the latter method is much less computationally expensive. Regarding satellites, it was observed that the performance is related to the satellite inter‐track distance, as higher number of observation sites allows more accurate bed elevation estimation.
Print ISSN:
0043-1397
Electronic ISSN:
1944-7973
Topics:
Architecture, Civil Engineering, Surveying
,
Geography
