Hydraulic parameter identification using satellite earth imagery Export

@article{citeulike:965125, abstract = {Despite of the progresses recently realized in the implementation of open-channel flow models, the determination of the parameters involved in the simulation process is still uncertain. In alternative to traditional measurements in the field, the use of high resolution satellite earth imagery (visible satellite, infrared, radar) is considered to ascertain, implementing optimization methods, the value of a set of hydraulic parameters allowing to characterize the flow with a precision sufficient to make flood studies. These satellite images generally give a top sight of the flow or of the flooded area. The scope of data assimilation is to make the best possible estimate of the state of a physical system, given data and a model describing the phenomenon. This study focuses only on sequential methods, that is to say methods that correct the model state at the moment of the observations. Data assimilation techniques can be divided into two classes according to the processes of resolution employed. Variational methods minimize a cost function that is the sum of a distance to the observations and a distance to an a priori estimate (often a prevision) of the model state. Statistical methods or filters explicitly solve the assimilation problem by calculating the linear optimal combination between guess and observations that minimizes the estimate error variance. The most known filter, the Best Linear Unbiased Estimator or B.L.U.E., has been proposed by Kalman in 1960. Both approaches have been tested on a simple case. Parameter identification procedure has been implemented for a mono-dimensional steady flow in compound channel with a trapezoidal main channel and near horizontal overbanks. The observations or gauged data, that could be made by a satellite, are created by adding a gaussian noise, inherent to the interpretation of satellite images, to flow top width. Flow top width is obtained by a 1D hydraulic simulation of Saint-Venant equations realized on a known river. The inverse problem is then implemented to identify the main characteristics of the waterway. The implemented methods and the results do not depend on the choice of the parameters to estimate. Even in lack of initial information, the estimated parameters allow to simulate the flow top width with an induced error of less than 1\%. The aim of these tests was to find the most suitable method for hydraulic parameter identification. This method is now been implemented on the Aisne river (North of France).}, author = {Roux, H. and Raclot, D. and Dartus, D. and Puech, C.}, citeulike-article-id = {965125}, citeulike-linkout-0 = {http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2003EAEJA.....1130R}, journal = {EGS - AGU - EUG Joint Assembly, Abstracts from the meeting held in Nice, France, 6 - 11 April 2003, abstract \#1130}, keywords = {assimilation, parameter, rs}, month = {April}, pages = {1130+}, posted-at = {2006-11-28 10:52:31}, priority = {2}, title = {Hydraulic parameter identification using satellite earth imagery}, url = {http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2003EAEJA.....1130R}, year = {2003} }

TY - JOUR ID - citeulike:965125 L3 - citeulike-article-id:965125 N2 - Despite of the progresses recently realized in the implementation of open-channel flow models, the determination of the parameters involved in the simulation process is still uncertain. In alternative to traditional measurements in the field, the use of high resolution satellite earth imagery (visible satellite, infrared, radar) is considered to ascertain, implementing optimization methods, the value of a set of hydraulic parameters allowing to characterize the flow with a precision sufficient to make flood studies. These satellite images generally give a top sight of the flow or of the flooded area. The scope of data assimilation is to make the best possible estimate of the state of a physical system, given data and a model describing the phenomenon. This study focuses only on sequential methods, that is to say methods that correct the model state at the moment of the observations. Data assimilation techniques can be divided into two classes according to the processes of resolution employed. Variational methods minimize a cost function that is the sum of a distance to the observations and a distance to an a priori estimate (often a prevision) of the model state. Statistical methods or filters explicitly solve the assimilation problem by calculating the linear optimal combination between guess and observations that minimizes the estimate error variance. The most known filter, the Best Linear Unbiased Estimator or B.L.U.E., has been proposed by Kalman in 1960. Both approaches have been tested on a simple case. Parameter identification procedure has been implemented for a mono-dimensional steady flow in compound channel with a trapezoidal main channel and near horizontal overbanks. The observations or gauged data, that could be made by a satellite, are created by adding a gaussian noise, inherent to the interpretation of satellite images, to flow top width. Flow top width is obtained by a 1D hydraulic simulation of Saint-Venant equations realized on a known river. The inverse problem is then implemented to identify the main characteristics of the waterway. The implemented methods and the results do not depend on the choice of the parameters to estimate. Even in lack of initial information, the estimated parameters allow to simulate the flow top width with an induced error of less than 1%. The aim of these tests was to find the most suitable method for hydraulic parameter identification. This method is now been implemented on the Aisne river (North of France). JF - EGS - AGU - EUG Joint Assembly, Abstracts from the meeting held in Nice, France, 6 - 11 April 2003, abstract \#1130 TI - Hydraulic parameter identification using satellite earth imagery SP - 1130 KW - assimilation KW - parameter KW - rs AU - Roux, H AU - Raclot, D AU - Dartus, D AU - Puech, C PY - 2003/April// UR - http://adsabs.harvard.edu/cgi-bin/nph-bib_query?bibcode=2003EAEJA.....1130R ER -