The traditional way to determine discharge in a stream or river is to measure water height and determine a stage-discharge curve by going out regularly with a water flow measurement device, like an Ott mill or an acoustic in- strument. The last part is costly because it takes a lot of time and difficult logistics. Moreover, another problem relies on the fact that since the geo- metry of the channel changes with each major storm, Measurements should be redone continuously for most natural streams. Also, the present equip- ment is expensive, not robust, and power-hungry. We would like to change this with the aids of data assimilation to build cost-effective stage-discharge curves. Some Data assimilation approach has been compared, and the Kal- man filter was chosen to be the most effective one for our purpose. During a dry period, the geometry of a streambed is measured with drones. Then, the water height and the velocity of the upper layer of the stream is measured by radar or camera. With the water level, we ran a hydraulic model, based on the Delft3D software. With a Kalman filter, the observed and modeled velocities are weighed, giving an optimal estimate of the velo- city and full discharge. The procedure would be repeated for both relatively wide estuary and rel- atively long river, and the result for both modelings would be evaluated and compared to each other. Kalman filter procedure would be applied to both uncorrelated Gaussian noise and spatially correlated noise. Moreover, the effects of perturbed observation with bias would be discussed. Consequently, by means of data assimilation, we can reach our goal, which is estimating streamflow in a cost-effective and relatively accurate way with the help of our observed data and hydraulic modeling. Finally, some tables would be presented as an indication of the effectiveness of the Data Assimila- tion procedure in order to predict the water height and surface water velocityfor both estuary and river.
Il modo tradizionale per determinare lo scarico in un torrente o fiume è misurare l'altezza dell'acqua e determinare una curva di scarico dello stadio uscendo regolarmente con un dispositivo di misurazione del flusso d'acqua, come un mulino Ott o uno strumento acustico. L'ultima parte è costosa perché richiede molto tempo e logistica difficile. Inoltre, un altro problema si basa sul fatto che, poiché la geometria del canale cambia con ogni tempesta maggiore, le misure dovrebbero essere ripetute continuamente per la maggior parte dei flussi naturali. Inoltre, l'attrezzatura attuale è costosa, non robusta e assetata di energia. Vorremmo cambiarlo con l'aiuto dell'assimilazione dei dati per costruire curve di scarica in fase convenienti. È stato confrontato un approccio di assimilazione dei dati e il filtro Kalman è stato scelto per essere il più efficace per il nostro scopo. Durante un periodo secco, la geometria di un flusso viene misurata con i droni. Quindi, l'altezza dell'acqua e la velocità dello strato superiore del flusso vengono misurate dal radar o dalla telecamera. Con il livello dell'acqua, abbiamo lanciato un modello idraulico, basato sul software Delft3D. Con un filtro Kalman, le velocità osservate e modellate vengono pesate, fornendo una stima ottimale della velo-città e della scarica completa. La procedura sarebbe ripetuta sia per un estuario relativamente ampio che per un fiume relativamente lungo, e il risultato per entrambi i modelli sarebbe valutato e confrontato tra loro. La procedura del filtro di Kalman verrebbe applicata sia al rumore gaussiano non correlato sia al rumore spazialmente correlato. Inoltre, verrebbero discussi gli effetti dell'osservazione perturbata con distorsione. Di conseguenza, attraverso l'assimilazione dei dati, possiamo raggiungere il nostro obiettivo, che è la stima del flusso di flusso in modo conveniente e relativamente accurato con l'aiuto dei nostri dati osservati e della modellazione idraulica. Infine, alcune tabelle verrebbero presentate come un'indicazione dell'efficacia della procedura di assimilazione dei dati al fine di prevedere l'altezza e la velocità delle acque superficiali sia per l'estuario che per il fiume.
Estimation of stream flow based on a combination of hydraulic modelling and data assimilation
RIZEHBANDI, AREF
2019/2020
Abstract
The traditional way to determine discharge in a stream or river is to measure water height and determine a stage-discharge curve by going out regularly with a water flow measurement device, like an Ott mill or an acoustic in- strument. The last part is costly because it takes a lot of time and difficult logistics. Moreover, another problem relies on the fact that since the geo- metry of the channel changes with each major storm, Measurements should be redone continuously for most natural streams. Also, the present equip- ment is expensive, not robust, and power-hungry. We would like to change this with the aids of data assimilation to build cost-effective stage-discharge curves. Some Data assimilation approach has been compared, and the Kal- man filter was chosen to be the most effective one for our purpose. During a dry period, the geometry of a streambed is measured with drones. Then, the water height and the velocity of the upper layer of the stream is measured by radar or camera. With the water level, we ran a hydraulic model, based on the Delft3D software. With a Kalman filter, the observed and modeled velocities are weighed, giving an optimal estimate of the velo- city and full discharge. The procedure would be repeated for both relatively wide estuary and rel- atively long river, and the result for both modelings would be evaluated and compared to each other. Kalman filter procedure would be applied to both uncorrelated Gaussian noise and spatially correlated noise. Moreover, the effects of perturbed observation with bias would be discussed. Consequently, by means of data assimilation, we can reach our goal, which is estimating streamflow in a cost-effective and relatively accurate way with the help of our observed data and hydraulic modeling. Finally, some tables would be presented as an indication of the effectiveness of the Data Assimila- tion procedure in order to predict the water height and surface water velocityfor both estuary and river.File | Dimensione | Formato | |
---|---|---|---|
aref_thesis (3).pdf
non accessibile
Dimensione
4.36 MB
Formato
Adobe PDF
|
4.36 MB | Adobe PDF | Visualizza/Apri |
I documenti in POLITesi sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
https://hdl.handle.net/10589/152782