 |
| Titel |
Identifiability of altimetry-based rating curve parameters in function of river morphological parameters |
| VerfasserIn |
Adrien Paris, Pierre André Garambois, Stéphane Calmant, Rodrigo Paiva, Collischonn Walter, Joecila Santos da Silva, Daniel Medeiros Moreira, Marie-Paule Bonnet, Frédérique Seyler, Jérôme Monnier |
| Konferenz |
EGU General Assembly 2016
|
| Medientyp |
Artikel
|
| Sprache |
en
|
| Digitales Dokument |
PDF |
| Erschienen |
In: GRA - Volume 18 (2016) |
| Datensatznummer |
250136976
|
| Publikation (Nr.) |
 EGU/EGU2016-18142.pdf |
|
|
|
|
|
| Zusammenfassung |
| Estimating river discharge for ungauged river reaches from satellite measurements is not
straightforward given the nonlinearity of flow behavior with respect to measurable and non
measurable hydraulic parameters. As a matter of facts, current satellite datasets do not give
access to key parameters such as river bed topography and roughness. A unique set of almost
one thousand altimetry-based rating curves was built by fit of ENVISAT and Jason-2 water
stages with discharges obtained from the MGB-IPH rainfall-runoff model in the Amazon
basin. These rated discharges were successfully validated towards simulated discharges (Ens
= 0.70) and in-situ discharges (Ens = 0.71) and are not mission-dependent. The rating
curve writes Q = a(Z-Z0)b*sqrt(S), with Z the water surface elevation and S its
slope gained from satellite altimetry, a and b power law coefficient and exponent
and Z0 the river bed elevation such as Q(Z0) = 0. For several river reaches in the
Amazon basin where ADCP measurements are available, the Z0 values are fairly well
validated with a relative error lower than 10%. The present contribution aims at
relating the identifiability and the physical meaning of a, b and Z0given various
hydraulic and geomorphologic conditions. Synthetic river bathymetries sampling a
wide range of rivers and inflow discharges are used to perform twin experiments. A
shallow water model is run for generating synthetic satellite observations, and then
rating curve parameters are determined for each river section thanks to a MCMC
algorithm.
Thanks to twin experiments, it is shown that rating curve formulation with water surface
slope, i.e. closer from Manning equation form, improves parameter identifiability. The
compensation between parameters is limited, especially for reaches with little water
surface variability. Rating curve parameters are analyzed for riffle and pools for
small to large rivers, different river slopes and cross section shapes. It is shown
that the river bed elevation Z0is systematically well identified with relative errors
on the order of a few %. Eventually, these altimetry-based rating curves provide
morphological parameters of river reaches that can be used as inputs into hydraulic
models and a priori information that could be useful for SWOT inversion algorithms. |
|
|
|
|
|
|