HAL will be down for maintenance from Friday, June 10 at 4pm through Monday, June 13 at 9am. More information
Skip to Main content Skip to Navigation

Sensibilité des calculs hydrologiques à la densité des réseaux de mesure hydrométrique et pluviométrique

Abstract : Rainfall and runoff data are very important to make hydrological computations for various objectives. Precipitation is essential information for water balance studies and for hydrological simulation and forecasting, since it is used as input of hydrological models. Runoff data are also essential to calibrate and validate models: they provide information on hydrological regimes, streamflow extremes, past trends and catchment hydrological behavior. Since precipitation and runoff are spatially and temporally variable, a good spatiotemporal representativeness of runoff and rainfall data is crucial to limit the uncertainty in hydrological computations. Therefore the presence of sufficiently dense hydrometeorological monitoring networks to account for this variability is essential. However, these networks may be expensive for their managers, leading to reflections on their rationalization. This rationalization, which often means a reduction of network density, can limit our knowledge of the hydrological cycle, and can significantly increase the uncertainties in hydrological computation. Quantifying this increase raises several difficulties, since it depends on the hydrological objectives, tools used and catchment characteristics. The main objective of this research was to study the impact of the spatiotemporal density of runoff and rainfall networks on the performances of several hydrological computations (namely simulation of daily streamflow, estimation of long term average streamflow and extreme streamflows). To get general results, studies were based on a large set of French catchments. The first part of this thesis focused on the impact of the hydrometric network density for ungauged or poorly gauged catchments. For ungauged catchments, the robustness of regionalization approaches was first analyzed by two methods of reduction of neighboring network density (hydrometrical desert and random reduction). Then the hydrometrical desert method was used to evaluate the sensitivity of hydrological computations on the spatial availability of runoff data. Our results suggest that for all the computations of streamflow considered here, the efficiency of the regionalization process decreases when the flow gauging network density is reduced. Results also show that this drop in efficiency is lower than the drop due to the regionalization method itself (i.e. going from gauged to ungauged situation; nothing beats the observed data on the studied sites). In a second step, we confirmed the value of using a few flow measurements on ungauged catchments, by combining this information with regional information. We pushed the analysis further by focusing on the differences between redundant measurements and random measurements, and by proposing equivalences. The second part focused on the impact of the rainfall network density on several hydrological computations. Results are less generalizable than those on the hydrometric network, revealing various trends within the set of catchments and between the hydrological computations considered. However, the decrease of GR4J model performances when the density of rainfall network is reduced seems to be related to the spatial variability of rainfall in the catchment.
Document type :
Complete list of metadata

Cited literature [245 references]  Display  Hide  Download

Contributor : Migration Irstea Publications Connect in order to contact the contributor
Submitted on : Saturday, May 16, 2020 - 7:19:45 AM
Last modification on : Monday, January 3, 2022 - 1:55:19 PM


Files produced by the author(s)


  • HAL Id : tel-02602066, version 1
  • IRSTEA : PUB00045673



L. Lebecherel. Sensibilité des calculs hydrologiques à la densité des réseaux de mesure hydrométrique et pluviométrique. Sciences de l'environnement. Doctorat Géosciences, Ressources Naturelles et Environnement, Spécialité : Hydrologie, Institut des Sciences et Industries du Vivant et de l'Environnement AgroParisTech, 2015. Français. ⟨tel-02602066⟩



Record views


Files downloads