Flood mapping

Floods are the most common and damaging natural hazard faced by many countries throughout the world. Understanding flood risk is of great importance to the management of socio-economic and environmental impacts.  

Several researchers produced alternative low-complexity solutions that rely on data-driven methods (Schumann et al., 2014a; Tang et al., 2018; Giovannettone et al., 2018; Caprario & Finotti, 2019; Zhao et al., 2019). Among them, hydrogeomorphic methods take advantage of the causality between historical floods and the floodplain hydraulic geometry (e.g., Bhowmik, 1984; McGlynn and Seibert, 2003; Dodov and Foufola-Georgiou, 2006) and make use of digital elevation models (DEMs) (see e.g. Tavares da Costa et al., 2019; Annis et al., 2019; Nardi et al., 2006, 2013, 2019; Clubb et al., 2017; Jafarzadegan and Merwade, 2017; De Risi et al., 2015; Degiorgis et al., 2012). 

The original idea that the hydrological responses of a catchment could be indexed on the basis of topography dates back to Kirkby (1975). He proposed the use of the TOPMODEL topographic index, which has proven to be a good indicator of flood-prone areas (Manfreda et al., 2011; De Risi et al., 2013; Jalayer et al., 2014). Its ability in reproducing the flood-prone areas was improved by Manfreda et al. (2011) who compared this ‘modified Topographic Index’ (TIm) with flood inundation maps obtained from hydraulic simulations, and observed that the portion of a basin exposed to flood inundation can be detected applying a threshold value, τ, on the map of TIm. This paper led to a number of applications where geomorphic indices, such as the TI and TIm, are adopted for the delineation of flood-prone areas ( e.g., Degiorgis et al., 2012; Suwandana et al., 2012; Knitter  et al., 2012; Jalayer et al., 2014; Cooper, 2014; Papaioannou et al., 2015; De Risi et al., 2014; Pourali et al., 2016). In some of these studies, the procedure developed by Manfreda et al. (2011) and the relative GRASS GIS tool (Di Leo et al., 2011) have been applied in its original form. In other studies, the procedure has been modified exploring new calibration strategies (e.g. Jalayer et al., 2014).

Apart from the use of the modified topographic index, the potential use of geomorphologic features for the delineation of flood prone area have been explored in a number of papers. Starting from simple information about basin features, these methods have been significantly improved with time, reaching good accuracy. 

In particular, Degiorgis et al. (2012) proposed the delineation of flood-prone areas from a location where a flood map exists to one where it does not. This extrapolation procedure was achieved by threshold binary classification or, in other words, identifying the isoline (optimal threshold, TH) of a chosen flood descriptor that best approximated the areal extent of an existing flood map. Flood descriptors can be defined as quantitative layers extracted from DEMs that correlate to the tendency of an area to flood.

Manfreda et al. (2014, 2015) and Samela et al. (2016) improved the method introduced by Degiorgis et al. (2012) by comparing and evaluating different descriptors in terms of their suitability to delineating flood-prone areas. In their studies, the Geomorphic Flood Index (GFI, Samela et al., 2017), was found to be the best performing and the most consistent hydrogeomorphic descriptor amongst the ones analysed (Manfreda et al., 2015; Samela et al., 2016, 2017), one of which was the Height Above the Nearest Drainage (HAND) (Rennó et al., 2008; Nobre et al., 2016). Building upon this, Samela et al. (2017) and Tavares da Costa et al. (2019a) successfully delineated flood-prone areas at the continental scale by dramatically reducing computational times and costs, opening new possibilities for flood risk assessment and management over large-scales. In Tavares da Costa et al. (2019a), optimal thresholds of the GFI were also shown to be positively correlated to flood extents associated with specific return periods.

Figure 1. Description of the two components that constitute the Geomorphic Flood Index (GFI) layer. a) Representation of a D8 flow direction raster for a portion of an elementary catchment. b) The empirically derived water level in each cell under analysis computed as a power law of bankfull depth and upslope contributing area of the hydrologically connected river centreline cell. c) The DEM elevation difference between the cell under analysis and the hydrologically connected river centreline cell. At the lower left, a river basin representation showing the elementary catchment E, in grey, and the river network, in blue. At the lower centre, a cross-section representation of the river channel and floodplain illustrating the two terms involved in the computation of the GFI for a generic flow path.
Figure 2. Schematic description of the threshold binary classification based on the Geomorphic Flood Index, GFI.
Figure 3. Panel 1 shows the flood-prone areas identified in the continental U.S. according to the linear binary classifier based on the Geomorphic Flood Index for a return period of 500-year (depicted in dark blue). The following two couples of images provide a more detailed plot of the flood-prone areas identified using this approach in two areas characterized by gaps in the FEMA floodplain map (turned into a binary map).

The intrinsic properties of the GFI have been further investigated by Manfreda and Samela (2019) to obtain an approximate, but immediate estimate of the water surface elevation in a river and surrounding areas. Surface elevations can be coupled with the vulnerability of exposed assets to obtain an assessment of the direct economic damage in a simple and effective way (Albano et al., 2020).

Figure 4. Estimation of the water depth based on the Geomorphic Flood Index (GFI) method. In particular, the map shows the comparison between the flood map derived by hydraulic simulations (panel a) and the flood map derived by the linear binary classifier based on the GFI (panel b). For both panels, the colour gradient represents inundation depths in meters, as reported in the legend.

Reference

Albertini, C., D. Miglino, V. Iacobellis, F. De Paola, S. Manfreda, Flood-prone areas delineation in coastal regions using the Geomorphic Flood IndexJournal of Flood Risk Management, e12766, 2021. [pdf]

Albano, R., Samela, C., Crăciun, I., Manfreda, S., Adamowski, J., Sole, A., … & Ozunu, A. Large Scale Flood Risk Mapping in Data Scarce Environments: An Application for RomaniaWater, 12(6), 1834, 2020. [pdf]

Tavares da Costa, R., S. Zanardo, S. Bagli, A. G. J. Hilberts, S. Manfreda, C. Samela, and A. Castellarin, Predictive modelling of envelope flood extents using geomorphic and climatic-hydrologic catchment characteristicsWater Resources Research, (doi: 10.1029/2019WR026453), 2020. [pdf]

Tavares da Costa, R., S. Manfreda, V. Luzzi,  C. Samela, P. Mazzoli, A. Castellarin, S. Bagli, A web application for hydrogeomorphic flood hazard mappingEnvironmental Modelling and Software, (doi: 10.1016/j.envsoft.2019.04.010) 2019.  [pdf]

Manfreda, S., C. Samela, A DEM-based method for a rapid estimation of flood inundation depthJournal of Flood Risk Management, 12 (Suppl. 1):e12541, (doi: 10.1111/jfr3.12541) 2019. [pdf]

Manfreda S., C. Samela, A. Refice, V. Tramutoli, F. Nardi, Advances in Large Scale Flood Monitoring and DetectionHydrology, 5, 49, (doi: 10.3390/hydrology5030049) 2018. [pdf]

Samela, C., R. Albano, A. Sole, S. Manfreda, A GIS tool for cost-effective delineation of flood-prone areasComputers, Environment and Urban Systems, 70, 43-52 (doi: 10.1016/j.compenvurbsys.2018.01.013), 2018.  [pdf]

Samela, C., T.J. Troy, S. Manfreda, Geomorphic classifiers for flood-prone areas delineation for data-scarce environmentsAdvances in Water Resources,  102, 13-28, (doi: 10.1016/j.advwatres.2017.01.007), 2017. [pdf]

Samela C., S. Manfreda, T. J. Troy, 100-year geomorphic flood-prone areas for the continental U.S., Data in Brief, 12, 203-207, (doi: 10.1016/j.dib.2017.03.044), 2017. [pdf]

D’Addabbo, A., A. Refice, G. Pasquariello, F. Lovergine, D. Capolongo and S. Manfreda, A Bayesian Network for Flood Detection Combining SAR Imagery and Ancillary DataIEEE Transactions on Geoscience and Remote Sensing, 54(6), 3612 -3625, (doi: 10.1109/TGRS.2016.2520487), 2016.  [pdf]

Samela, C., S. Manfreda, F. De Paola, M. Giugni, A. Sole, M. Fiorentino, DEM-based approaches for the delineation of flood prone areas in an ungauged basin in AfricaJournal of Hydrologic Engineering, 21(2), (doi: 10.1061/(ASCE)HE.1943-5584.0001272), 2016.  [pdf

Manfreda, S., C. Samela, A. Gioia, G. Consoli, V. Iacobellis, L. Giuzio, A. Cantisani, A. Sole, Flood-Prone Areas Assessment Using Linear Binary Classifiers based on flood maps obtained from 1D and 2D hydraulic modelsNatural Hazards, 79 (2), 735-754, (doi: 10.1007/s11069-015-1869-5), 2015. [pdf]

Manfreda, S., F. Nardi, C. Samela, S. Grimaldi, A. C. Taramasso, G. Roth and A. Sole, Investigation on the Use of Geomorphic Approaches for the Delineation of Flood Prone AreasJournal of Hydrology, 517, 863-876, (doi: ), 2014. [pdf]

Manfreda, S. and Sole, A. Closure to “Detection of Flood-Prone Areas Using Digital Elevation Models” by Salvatore Manfreda, Margherita Di Leo, and Aurelia SoleJournal of Hydrologic Engineering, 18(3), 362-365, (doi: 10.1061/(ASCE)HE.1943-5584.0000693), 2013.  [pdf]

Manfreda, S., M. Di Leo, A. Sole, Detection of Flood Prone Areas using Digital Elevation ModelsJournal of Hydrologic Engineering, 16(10), 781-790  (doi: 10.1061/(ASCE)HE.1943-5584.0000367), 2011.  [pdf]

Attachments