The effectiveness of the CN method in areas with saturated soil conditions

Abstract

The Soil Conservation Service (SCS) Curve Number (CN) method has been widely used by engineers in hydrological modelling. The main advantage of this method is that only a few parameters are needed to calculate the CN. The purpose of this study was to analyse the effectiveness of the regionalized CN method for the analysis of flood vulnerability in saturated soils, commonly located in coastal floodplains. The study area is located in the Juqueriquere River Basin, on the northern coastline of Sao Paulo, Brazil. It holds the major non-urbanised plains of the northern coastline of the state. It is constrained by the high altitude mountains of Serra do Mar, which causes intense orographic rain in the area. Added to the basin’s geophysical features, the influence of tide contributes to the local high vulnerability to floods. Even though the area is not representatively gauged, and scarce runoff data is available, there is high interest to urbanise the plains, due to the recent implantation of the Gas Treatment Unit of Caraguatatuba (UTGCA) of the Brazilian Petroleum Corporation (PETROBRAS) in 2012, and the proximity to the Port of Sao Sebastiao. The City Master Plan of Caraguatatuba Municipality (CMPC) was proposed in 2011. Both the gas pipelines of UTGCA and the part of the Tamoios Highway Complex have been implemented, based on the Environmental Impact Assessment of PETROBRAS (2007), but no macro or micro drainage plans have been developed for this urbanizing area yet. The Foundation of Water Resources of Sao Paulo (Costa Norte 2017) has just approved the macro drainage plan for the downstream area, which is densely urbanised and constantly affected by floods. In the study, the runoff in each sub-basin was derived from the CN method at the Hydrologic Modeling System (HEC-HMS). The Manning's roughness coefficients of different cross-sections were calibrated at the Hydrologic Engineering Center's River Analysis System (HEC-RAS), where the CN method was also adopted. The rain gauge of PETROBRAS (EMQAR1), near UTGCA plant, recorded 247.20 mm in a 24h period between 17 and 18 March 2013. This event caused severe floods in the basin, especially in the upstream area. Thus, this rainfall event was used for calibration purposes, taking into consideration the water level of three different cross-sections registered by local farmers. The CN values of the present scenario were attributed by the land use and land cover (LULC) classification of high-resolution imagery, and were regionalised per sub-basin using the LULC area weighted average approach. The future scenario was based on the collection of similar perviousness-patterned zones of the proposed CMPC. Both scenarios were simulated with the same SCS unit hydrograph, regarding the same calibration event. The findings of the study revealed that, even though the peak discharge of the future scenario had higher values than the present scenario, they did not represent the significant increase of imperviousness of some of the CMPC zones. It occurred due to: the CN regionalization approach, potential CN losses and underestimation of the rainfall intensity, as the initial abstraction is not related to the soil infiltration.