GFA - tool is an open-source software to realize a fast and cost-effective delineation of the floodplains in the contexts where the available data is scarce to carry out hydrological/hydraulic analyses. This algorithm uses the linear binary classification technique based on the recently proposed Geomorphic Flood Index (GFI) (Samela et al., 2017).
Input:
DEM: Digital Elevation Model grid
Filled DEM: Depressionless DEM grid
Flow Direction D8: eight direction flow model grid
Flow Accumulation: Flow accumulation grid
Set methodology options:
FlowDir coding: Flow direction grid encoding (according to ESRI (1-128), or to HyGrid2k2 (0-315) or to TauDEM (1-8))
Drainage network identification method:
(i) ASk method: (Giannoni et al., 2005) the channel starts from location where the quantity AS^k (A=contributing area, S=local slope, k=drainage basin related exponent) is higher than a specific threshold;
(ii) FAt method: (Tarboton et al., 1991) the channel starts from location where the flow accumulation is higher than a specific threshold
Drainage network identification threshold: threshold value to identify channel network
Set calibration options:
Manually set threshold: set manually the threshold of GFI to distinguish flood prone areas and areas not prone to floods
GFI classifier threshold: set the value of the GFI threshold
Calibrate Threshold: automatic calibration of the GFI threshold (ROC curve method). A standard flood hazard map derived by hydraulic models, must be selected as the “gold standard truth” and used to train a linear binary classifier based on the GFI. This calibration map is necessary for limited portions of the basin of interest (at least 2% of drainage basin)
Output:
GFI raster: the Geomorphic Flood Index map (Samela et al., 2017)
GFI normalized raster: the Geomorphic Flood Index map, with values normalized in the range -1:1
GFI derived flood-prone areas map: the binary raster of the flood prone areas identified by the GFI classifier
GFI performance metrics: a text file which stores the calibration threshold, the false positive rate, the false negative rate, the area under the ROC curve
Create intermediate files: locally store all the main intermediate file utilized to produce the above results
References:
Giannoni, F., Roth, & G., Rudari, R., (2005). A procedure for drainage network identification from geomorphology and its application to the prediction of the hydrologic response. Advances in Water Resources, 28(6), 567-581.
Tarboton D. G., R. L. Bras, I. Rodriguez–Iturbe. 1991. On the Extraction of Channel Networks from Digital Elevation Data. Hydrological Processes. 5: 81–100.
Samela, C., Troy, T.J., & Manfreda, S. (2017). Geomorphic classifiers for flood-prone areas delineation for data-scarce environments, Advances in Water Resources 102: 13-28
Samela, C., Albano, R., Sole, A., Manfreda, S. (2018). Geomorphic Flood Area (GFA): a QGIS tool for a cost-effective delineation of the flood-prone areas, Computers, Environment and Urban Systems
