the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
| 13 May 2024
A multivariate statistical framework for mixed populations in compound flood analysis
Abstract. In coastal regions, compound flooding can arise from a combination of different drivers such as storm surges, high tides, excess river discharge, and rainfall. Compound flood potential is often assessed by quantifying the dependence and joint probabilities of the flood drivers using multivariate models. However, most of these studies assume that all extreme events originate from a single population. This assumption may not be valid for regions where flooding can arise from different generation processes, e.g., tropical cyclones (TCs) and extratropical cyclones (ETCs). Here we present a flexible copula-based statistical framework to assess compound flood potential from multiple flood drivers while explicitly accounting for different storm types. The proposed framework is applied to Gloucester City, New Jersey, and St. Petersburg, Florida as case studies. Our results highlight the importance of characterizing the contributions from TCs and non-TCs separately to avoid potential underestimation of the compound flood potential. In both study regions, TCs modulate the tails of the joint distributions (events with higher return periods) while non-TC events have a strong effect on events with low to moderate joint return periods. We show that relying solely on TCs may be inadequate when estimating compound flood risk in coastal catchments that are also exposed to other storm types. We also assess the impact of non-classified storms that are neither linked to TCs or ETCs in the region (such as locally generated convective rainfall events and remotely forced storm surges). The presented study utilizes historical data and analyzes two populations, but the framework is flexible and can be extended to account for additional storm types (e.g., storms with certain tracks or other characteristics) or can be used with model output data including hindcasts or future projections.
- Preprint
(12140 KB) - Metadata XML
-
Supplement
(899 KB) - BibTeX
- EndNote
Status: open (until 24 Jun 2024)
-
RC1: 'Comment on egusphere-2024-1122', Anonymous Referee #1, 31 May 2024
reply
This paper discusses making use of copulas and extreme value distributions conditional on different storm types to quantify annual exceedance probabilities of joint events at multiple locations, with a focus on two case studies in the USA. Events are described as being driven by non-tidal residuals or by rainfall caused by, in part, tropical and extratropical cyclones. The paper looks to analyse the relationships between flooding impact and these different drivers based on meteorological observations (ERA5) and hydrological modelling (UTide). Extreme events were identified using a Peaks-over-Threshold approach with a simple Independence/declustering criterion, and attributed to cyclones with a simple spatio-temporal proximity metric. Joint AEPs are based on "AND" scenarios (occurrence of events at all locations at the same time) and use a variety of copula families to determine these.Overall, looking at the two similar problems of event driver attribution and simultaneous events is fairly novel. Although the constituent parts are known standards, the combination is not one this reviewer has come across before, which is much to the paper's benefit. At times, this paper is spinning a lot of plates at once, but ultimately the conclusions are clearly shown by the methods and data. The stated flexibility to more than two sites simultaneously is mentioned but not demonstrated.Overall, this is a strong paper, which needs some adjustments to improve the communication of the methods and the conclusions.Major issues:
- The POT event declustering is quite simplistic and doesn't count for the varying speed of response of the flow to rainfall at each location.
- The reference to "high" and "low" return periods from line 330 onwards needs to be more specific. The exact choice of cutoff is not important, but stating what it is and sticking to it is important.
- Using high numbers of different copulas and distributions can make it hard to compare results, especially between sites. Using more general distributions (e.g. kappa) or more general copulas and sticking to one overall best choice may make it easier to interpret.
Minor issues:- The heavy use of abbreviations does make this harder to read, consider sometimes switching back to fully spelling out terms like rainfall.
- Check your references in the main body match those in the bibliography.
- Some paragraphs could be swapped for tables (paragraph around line 168, paragraph around line 330
- line 252: numbered lists should be presented as such, with a new line for each for readability.
- Figure 4: It is unclear what is exactly meant by "natural variability", and how it differs from the significance shown by the red circles.
- Figures 7 and 8 are very busy. Consider splitting into more figures, or remove elements of the figures which do not directly contribute to your conclusions.
Citation: https://doi.org/10.5194/egusphere-2024-1122-RC1 -
RC2: 'Comment on egusphere-2024-1122', Anonymous Referee #2, 08 Jun 2024
reply
This paper is well-written and scientifically sound. I would like to see more discussion on the following topics:
[1] you end up with results for RF and NTR. Please recommend how to combine these results in a flood risk analysis
[2] the isolines method has some shortcomings in my view. The paper hints that these can be used to define a single event (the one on the isoline with the highest probability density), However, in most compound flood risk analysis there is no single "most representative event". In coastal zone there are locations close to the coastline for which the main flood driver is the peak sea level (surge) and there are locations more inland for which the rainfall is the dominant flood river. For the former, an event with extreme NTR and moderate RF is the most relevant, for the latter an event with extreme RF and moderate NTR is the most relevant.
furthermore some minor comments in the attached document
Viewed
| HTML | XML | Total | Supplement | BibTeX | EndNote | |
|---|---|---|---|---|---|---|
| 200 | 93 | 8 | 301 | 22 | 7 | 9 |
- HTML: 200
- PDF: 93
- XML: 8
- Total: 301
- Supplement: 22
- BibTeX: 7
- EndNote: 9
Viewed (geographical distribution)
| Country | # | Views | % |
|---|
| Total: | 0 |
| HTML: | 0 |
| PDF: | 0 |
| XML: | 0 |
- 1