Are Feature Agreement Statistics Alone Sufficient to Validate Modelled Flood Extent Quality? A Study on Three Swedish Rivers Using Different Digital Elevation Model Resolutions

Abstract

Hydraulic modelling is now, at increasing rates, used all over the world to provide flood risk maps for spatial planning, flood insurance, etc. This puts heavy pressure on the modellers and analysts to not only produce the maps but also information on the accuracy and uncertainty of these maps. A common means to deliver this is through performance measures or feature statistics. These look at the global agreement between the modelled flood area and the reference flood that is used. Previous studies have shown that the feature agreement statistics do not differ much between models that have been based on digital elevation models (DEMs) of different resolutions, which is somewhat surprising since most researchers agree that high-resolution DEMs are to be preferred over poor resolution DEMs. Hence, the aim of this study was to look into how and under which conditions the different feature agreement statistics differ, in order to see when the full potential of high-resolution DEMs can be utilised. The results show that although poor resolution DEMs might produce high feature agreement scores (around F > 0.80), they may fail to provide good flood extent estimations locally, particularly when the terrain is flat. Therefore, when high-resolution DEMs (1 to 5 m) are used, it is important to carefully calibrate the models by the use of the roughness parameter. Furthermore, to get better estimates on the accuracy of the models, other performance measures such as distance disparities should be considered.

InundatEd: A Large-scale Flood Risk Modeling System on a Big-data – Discrete Global Grid System Framework

“…As noted in Lim & Brandt (2019), the reliability of the observed flood extent polygons also merits comment. In this case study, the observed FEPs for the Ottawa River watershed were originally digitized from remotely sensed data and thus carry forward the errors and uncertainties from prior processing.…”

Sentinel‐1 remote sensing data and Hydrologic Engineering Centres River Analysis System two‐dimensional integration for flash flood detection and modelling in New Cairo City, Egypt

“…To determine and compare how well the raster layers of the flood prone area obtained by SAR images fit with the raster of flooded areas simulated by HEC‐RAS 2D, two scales, feature agreement statistics F 1 and F 2 are used as shown in Equations (1) and (2) (Bates & De Roo, 2000; Lim & Brandt, 2019b; Mason, Bates, & Dall'Amico, 2009): where A is the total number of pixels showing a perfect match between SAR and the simulated model, B is false or overestimated matches, and C is the missing or underestimated pixels. More details about different verification scales for raster and vector data can found in Hunter (2005)and Lim and Brandt (2019a). The F values range from 0 to 100%, where 100% is the maximum performance.…”

“…where A is the total number of pixels showing a perfect match between SAR and the simulated model, B is false or overestimated matches, and C is the missing or underestimated pixels. More details about different verification scales for raster and vector data can found in Hunter (2005)and Lim and Brandt (2019a). The F values range from 0 to 100%, where 100% is the maximum performance.…”

InundatEd-v1.0: a height above nearest drainage (HAND)-based flood risk modeling system using a discrete global grid system

“…Afshari (2018) achieved F 1 values from 0.48 to 0.64 for the 10-year, 100-year, and 500-year return periods when comparing a HAND-based simulation against a Hydrologic Engineering Center's River Analysis System (HEC-RAS) 2-D control. Lim and Brandt (2019) determined that low-resolution DEMs are capable of yielding relatively high comparison metrics (e.g., F 1 values approximately >= 0.80) in situations where Manning's n varies widely over space. The connection between high values of Manning's n and flood overestimation (false discovery) was also discussed.…”

“…Platforms through which the public can view and interact with the flood extent maps may also be developed (Tavares da Costa et al, 2019). One such simple conceptual inundation model is the flood model based on height above nearest drainage (HAND) (Liu et al, 2018). Zheng et al (2018) estimated the river channel geometry and rating curve estimation using HAND, which gained interest from the community, industry, and government agencies.…”