Assessment of Flood Susceptibility Using Support Vector Machine in the Belt and Road Region
Abstract. Floods have occurred frequently all over the world. During 2000–2020, nearly half (44.9 %) of global floods occurred in the Belt and Road region because of its complex geology, topography, and climate. However, the degree of flood susceptibility of each sub-region and country in the Belt and Road region remains unclear. Here, based on 11 flood condition factors, the support vector machine (SVM) model was used to generate a flood susceptibility map. Then, we introduced the flood susceptibility comprehensive index (FSCI) for the first time to quantify the flood susceptibility levels of the sub-regions and countries in the Belt and Road region. The results reveal the following. (1) The SVM model used in this study has an excellent accuracy, and the AUC values of the success-rate curve and prediction-rate curve were higher than 0.9 (0.917 and 0.934 respectively). (2) The areas with the highest and high flood susceptibility account for 12.22 % and 9.57 % of the total study area respectively, and these areas are mainly located in the southeastern part of Eastern Asia, almost the entirely of Southeast Asia and South Asia. (3) Of the seven sub-regions in the Belt and Road region, Southeast Asia is most susceptible to flooding and has the highest FSCI (4.49), followed by South Asia. (4) Of the 66 countries in this region, 16 of the countries have the highest flood susceptibility level (normalized FSCI > 0.8) and 5 countries (normalized FSCI > 0.6) have a high flood susceptibility level. These countries need to pay more attention to flood mitigation and management. The above findings provide useful information for decision-making in flood management in the Belt and Road region. In the future study, higher quality flood points, and climate change factors should be considered.
Jun Liu et al.
Jun Liu et al.
Jun Liu et al.
Viewed (geographical distribution)
The authors present a data-driven model to predict flood susceptibility at ~10 km resolution, using 11 geographic datasets trained against point observations of flooding from DFO and EM-DAT. It is regrettable that I must judge this work to be of poor quality in many respects, and can therefore only recommend it be rejected for publication in NHESS.
Flood observations of this provenance over 20 years are not good enough to be skillful in the characterisation of flood hazard at this scale. Just because a flood hasn't been recorded since the year 2000, doesn't mean a flood did not happen (within the time period, or indeed before or after it). All your model has therefore done is replicate where DFO/EM-DAT has recorded floods in the past 20 years, having been trained on where DFO/EM-DAT has recorded floods in the past 20 years. The skill score (AUC) is extremely high, but it is skillfully replicating something that is neither useful nor interesting: GIGO. It is therefore not clear how any of the data produced in this study advance our understanding of flood hazard, or present novel methods capable of doing so.
The writing is generally of poor quality – for which I am sympathetic to the authors – but it does make the manuscript difficult and/or unpleasant to read in some places. The paper is too long, containing much superflueous information, much of which is incorrect anyway (for instance, claiming the GDP ($2.3bn) of a region spanning 3 continents being roughly equivalent to that of a small English market town). It is overloaded with references that are not needed, which are often unrelated to the point being made. I personally have never heard of the "Belt and Road region", and so would refer to the study area as something else (or make it less arbitrary). Indeed, there's no reason not to deploy this method globally, given its simplicity.
I am sorry to not be able to give a more positive review of this work.