Comparison of catchment and network delineation approaches in complex suburban environments: application to the Chaudanne catchment, France

Abstract

Suburban areas are subject to strong anthropogenic modifications, which can influence hydrological processes. Sewer systems, ditches, sewer overflow devices and retention basins are introduced and large surface areas are sealed off. The knowledge of accurate flow paths and watershed boundaries in these suburban areas is important for storm water management, hydrological modelling and hydrological data analysis. This study proposes a new method for the determination of the drainage network based on time efficient field investigations and integration of sewer system maps into the drainage network for small catchments of up to 10 km 2 . A new method is also proposed for the delineation of subcatchments and thus the catchment area. The subcatchments are delineated using a combination of an object-oriented approach in the urban zone and geographical information system-based terrain analysis with flow direction forcing in the rural zone. The method is applied to the Chaudanne catchment, which belongs to the Yzeron river network and is located in the suburban area of Lyon, France. The resulting subcatchment map gives information about subcatchment response and contribution. The method is compared with six other automatic catchment delineation methods based on stream burning, flow direction forcing and calculation of subcatchments for inlet points. None of the automatic methods could correctly represent the catchment area and flow paths observed in the field. The watershed area calculated with these methods differs by as much as 25% from the area computed with the new method.Method 1 is the proposed method, method 2 uses flow direction forcing, method 3 contains subcatchments for inlet points and methods 4a-4d use different kind of stream burning: method 4a uses just variable road burning, and methods 4b-4d use variable road burning and the burning of the drainage network of 1, 5 and 10 m, respectively. 3756 S. JANKOWFSKY ET AL.

Parameterization of a Hydrological Model for a Large, Ungauged Urban Catchment

“…The flow of stormwater in urban areas does not necessarily follow the topography due to obstacles such as buildings or street curbs [32]. Jankowfsky et al [33] showed for a peri-urban catchment in France that catchment properties (such as the drainage area) differed significantly when either only a DEM was used for delineation or the topographical information was complemented with field survey data. They stated, however, that for large parts of the catchment only a DEM with a low spatial resolution (25 m) was available.…”

Water displacement by sewer infrastructure in the Grote Nete catchment, Belgium, and its hydrological regime effects

“…Consequently the associated sewer infrastructure might transfer water between different natural (sub)catchments and further affect the natural hydrological processes in the catchment (Simmons and Reynolds, 1982). Recent research on catchment delineation considered incorporation of these changes in hydrological flow paths using semi-automated procedures (Jankowfsky et al, 2013). Such delineations, however, remain largely data dependent and time consuming.…”

“…After calibration of the large set of parameters in such models, a better match between simulated and observed hydrological variables may be obtained, but this does not necessarily mean that the model has a good accuracy. Model over parameterization and related parameter identifiability problems are well-known pitfalls (Beven, 1989;Jakeman and Letcher, 2003;Muleta and Nicklow, 2005). These problems limit the applicability of such models.…”

Land cover mapping using aerial and VHR satellite images for distributed hydrological modelling of periurban catchments: Application to the Yzeron catchment (Lyon, France)

“…The periurban space is composed of small objects, with small areas such as individual 23 houses, narrow roads, hedges... Their identification requires the use of VHR (Very High 24 Resolution) images. This spatial constraint led us to choose images taken from optical 1 sensors which presently offer the highest resolutions.…”

“…The use of these attributes allowed solving 20 the spectral confusions between for example, porous mineral areas (plowed) and 21 impervious mineral areas (artificial areas), the urban objects being generally of smaller 22 size as compared to farming land. 23 The recording date of the image played an important role. It impacted the development 24 stage of the vegetation and especially the cover fraction of herbaceous vegetation, 25 24 which grows from bare soil to low and then high chlorophyll content vegetation.…”