Analyse et modélisation multifractales des interactions ondes-turbulence-biologie dans un lac urbain

Abstract : Research in limnology has generally focused on natural lakes and dams. Moreover, the scientific challenges posed by urban lakes are numerous because of their large width and shallow depth resulting in complex problems that involve. This work is part of the project Petits Lacs Urbains Mesures Modèles Multi-Echelles (PLUMMME), founded by Région Île-de-France (DIM R2DS program). The project provides the high-resolution equipment measuring equipment used at Lake Créteil. Physical and biological measurements are the primary scope of this thesis, with some comparisons of measurements done between Lake Creteil and Lake Bourget. With this in mind, the first step of the thesis was to enhance currently available data. In order to do this we installed a station for the continuous measurement of: temperature, chlorophyll and light, also meteorological quantities such as wind speed, and air temperature were measured. The quantities were measured using two measuring chains comprising different sensors. Measurements were conducted in order to characterise hydrodynamics, using a current meter of type Acoustic Doppler Velocimeter (ADV) and a profiler of type Acoustic Doppler Current Profilers (ADCP) positioned at different points in the lake (centrally, and at the point of stormwater discharge). The analysis of the various measured fields shows that the lake is periodically stratified with a period of one week. We also identify the different modes of the internal waves that occur due to the effects of the wind. Spectral analysis was used to highlight the first scale invariant propriety of the different measured fields. Various physical processes (turbulence, stratification, near-wall flow) operating along the water column were also characterised using this method. The spectral analysis, however, does not provide information about the intermittency of the fluctuations of the measured fields, this can only be done using multifractal techniques. In this thesis, we have shown that when there is a dependence between two fields, the dependence is multiscale. The use of the Universal Multifractal (UM) model, allows one to quantify the degree of this dependence. The quality of the estimation of the UM parameters depends strongly on that of the observed scale: time-series with trends are not scaling. The effects of the latter on the estimation of UM parameters can be improved by the application of the Empirical Modal Decomposition method. The measured velocity data from the ADCP shows that the slope of the power spectra density follow a logarithmic profile along the depth of the lake, it shows that different physical processes operate along the water column. We also show that the hydrodynamics of the lake at small scale are strongly perturbed at the point of waterstorm discharge. The last part of the thesis is focused on what our analysis at small-scales brings to numerical models. We show that, if the deterministic models are able to reproduce some phenomena at large scale, they fail to describe the small-scale variability. the small scale variability and the physical processes involved. A multifractal analysis showed that the small-scale variability of the physical fields displays a strong intermittency, an extremely important feature for biological or chemical reactions and therefore for biological scenarios. Knowing that the majority of biological/chemical interactions occur at smaller scales, this result underlines the necessity to greatly improve the closure of the Navier-Stokes equations. Finally, we show that the structure function, a frequently used statistical tool in turbulence, do not uniquely characterize non-conservative fields, i.e., they do not correspond to identical simulations
Complete list of metadatas

Cited literature [177 references]  Display  Hide  Download

https://pastel.archives-ouvertes.fr/tel-01164689
Contributor : Abes Star <>
Submitted on : Wednesday, June 17, 2015 - 3:22:09 PM
Last modification on : Friday, October 4, 2019 - 1:29:07 AM
Long-term archiving on : Tuesday, April 25, 2017 - 11:13:47 AM

File

TH2014PEST1166_complete.pdf
Version validated by the jury (STAR)

Identifiers

  • HAL Id : tel-01164689, version 1

Citation

Yacine Mezemate. Analyse et modélisation multifractales des interactions ondes-turbulence-biologie dans un lac urbain. Sciences de la Terre. Université Paris-Est, 2014. Français. ⟨NNT : 2014PEST1166⟩. ⟨tel-01164689⟩

Share

Metrics

Record views

1979

Files downloads

370