10th International Conference on Geomorphology
© Author(s) 2022. This work is distributed under
the Creative Commons Attribution 4.0 License.

High resolution modelling of floodplain water storage capacity loss in a highly urbanized tropical catchment and its implications for flood management.

Iury Simas1, Cleide Rodrigues1, and Janet Hooke2
Iury Simas et al.
  • 1Departamento de Geografia - FFLCH, Universidade de São Paulo, São Paulo, Brazil (iurysimas@usp.br; cleidrig@usp.br)
  • 2School of Environmental Sciences, University of Liverpool, Liverpool, United Kingdom (hookej@liverpool.ac.uk)

The city of São Paulo in Brazil is the largest urban agglomeration in the world outside Asia. The urbanization process has developed extensively and with a high density of construction in areas of high drainage density of this tropical environment, including over natural floodplains and low terraces, which has made flooding the most prevalent type of risk. Historically, the containment of such events involves the implementation of artificial channels and the construction of artificial storage structures, which have proven to be insufficient, since the episodes of floods with great impacts on the population are repeated annually. Apart from intense tropical rains and increasing impermeabilization, studies suggest that the loss of natural flood-dampening capacity exerted by floodplains is critical to the extent of flood events. In order to evaluate this factor in the occurrence of floods observed in recent years, a highly urbanized sample basin was selected and high-resolution airborne Lidar data was obtained. Differentiating the points, it was possible to classify the surface elements, isolating the terrain, vegetation, buildings and suspended structures such as bridges and transmission lines. The delimitation of the floodplain’s natural extension was carried out through stereoscopic restitution, using aerial photos from the pre-urbanization period, also supported by the terrain model obtained by classified Lidar. As a result, a GIS terrain modeling stage was carried out, in which different water level heights were tested to completely fill the floodplain. This process was undertaken by subdividing the original floodplain surface into blocks, so that this hypothetical water surface could be adjusted to the existing slope between the different sections. Obtaining the volume of water needed to completely fill the natural floodplain, it was then possible to obtain the volume occupied by buildings within this simulated area. Likewise, it was possible to run the same simulation for floods with different heights, from 0.5 to 6.5 meters above the bankfull stage (value identified as the threshold for complete filling of the floodplain). This analysis showed that the average storage volume lost due to the presence of buildings decreases as the water level is raised in the simulated events (in terms of cubic meters lost per meter of water column). This allows us to identify that one of the most immediate effects of the presence of buildings on the floodplain is the widening of floodspreads in events of lesser magnitude. When testing the complete filling of the floodplain, the volume occupied by the buildings corresponds to 25% of the volume of water that could be stored by its natural morphology. This volume of lost storage exceeds by 22 times the capacity planned by public authorities for the installation of artificial storage structures. Furthermore, considering a recent high-magnitude flood event (5 meters above bankfull stage) the lost storage volume exceeds the projected capacity by 17 times. Thus, it is considered that the structural engineering measures, already applied and planned, would be insufficient and impractical to compensate the loss of this environmental service naturally provided by the floodplain.

How to cite: Simas, I., Rodrigues, C., and Hooke, J.: High resolution modelling of floodplain water storage capacity loss in a highly urbanized tropical catchment and its implications for flood management., 10th International Conference on Geomorphology, Coimbra, Portugal, 12–16 Sep 2022, ICG2022-294, https://doi.org/10.5194/icg2022-294, 2022.