The Return Period Analysis of Heavy Rainfall Disasters Based on Copula Joint Statistical Modeling

Liu, Siyu; Dong, Xuguang

doi:https://doi.org/10.5194/egusphere-2024-2139

Preprints

https://doi.org/10.5194/egusphere-2024-2139

Preprints

09 Sep 2024

| 09 Sep 2024

Status: this preprint is open for discussion.

The Return Period Analysis of Heavy Rainfall Disasters Based on Copula Joint Statistical Modeling

Siyu Liu and Xuguang Dong

Abstract. In the last few years, with the frequent occurrence of extreme weather across the globe, it has become clear that a comprehensive understanding of the patterns and main characteristics of disaster occurrence is essential, and the willingness to study these variables has become more urgent than ever. This paper analyses the multivariate and spatial distribution characteristics of heavy precipitation disasters and proposes a method for estimating the degree of disaster-causing risk using a joint statistical model. This paper tests the model's validity with hourly precipitation data from 122 national meteorological stations in Shandong from 1990 to 2023. Based on heavy precipitation events in the past thirty years, different marginal distribution functions fit the duration of heavy precipitation and precipitation amount. The joint probability distribution model of two related variables is established based on the Copula joint distribution to analyze the change rule of heavy precipitation recurrence period in different periods and to analyze the characteristics of heavy precipitation causing disasters in Shandong Province on this basis. Compared with the disaster return period calculated by relying on univariate variables, the Copula function can more reasonably simulate the natural occurrence of the degree of disaster. The joint return period (JRP) estimated by the Copula function shows that the JPR of heavy rainfall with a duration of 1 hour is 89 % higher than that of 6 hours, indicating a significant increase in the risk of disasters caused by short-term heavy rainfall in Shandong region. This method can more scientifically describe the risk of disasters caused by heavy precipitation in different scenarios, especially the characteristics of disasters caused by short-term heavy precipitation, which can provide an adequate scientific basis for disaster prevention and mitigation planning and disaster risk management.

Received: 10 Jul 2024 – Discussion started: 09 Sep 2024

Publisher's note: Copernicus Publications remains neutral with regard to jurisdictional claims made in the text, published maps, institutional affiliations, or any other geographical representation in this preprint. The responsibility to include appropriate place names lies with the authors.

Download & links

Siyu Liu and Xuguang Dong

Status: open (until 24 Oct 2024)

Post a comment Subscribe to comment alert

RC1: 'Comment on egusphere-2024-2139', Anonymous Referee #1, 09 Oct 2024 reply

Title: The Return Period Analysis of Heavy Rainfall Disasters Based on Copula Joint Statistical Modeling
This manuscript intends to use a new statistical approach to evaluate past heavy rainfall events and return periods, aiming to provide a new tool for decision-makers and urban planning. My general comments can be found as follows:

If the authors' main idea is not a global approach, they must specify the study region location in the title of the manuscript.
Abstract:
Lines 15-16: The forecast of extreme weather, especially regarding extreme rainfall events, is a global challenge due to intrinsic physical and local environment dependence, which can originate from it. So, the authors must be cautious in proposing a statistical approach to forecast (estimate) them.
Line 18: Specify the country where Shandong is located.
Lines 32-34: How can a statistical approach for past events be used to forecast possible future extreme rainfall events?
Line 59: Heavy precipitation and extreme rainfall precipitation are distinct terms. Please clarify and use an adequate definition throughout the manuscript.
Lines 53-57: The authors mention the number related to the damage caused by an extreme rainfall event but do not specify the precipitation volume recorded. Please insert this information.
Line 60: Return period, also known as return time, recurrence interval, or recurrence time, is the estimated interval between occurrences of equal magnitude of a natural phenomenon. Please correct the definition.
Line 63: Please specify the "previous studies".
Line 79: Specify the "flexible assumptions" made.
Lines 92-93: The authors explain that the Copula function could be an alternative to the classical statistical approach for considering the variables' independence. However, the work of De Michele et al. (2023) used variables that are interligated. How do they explain that?
Lines 103-104: Rewrite the sentence. It isn't apparent.
Line 105: What problems do the authors intend to resolve?
Lines 105-130: The paragraph is confusing and complicated to understand. Please rewrite.
Line 154-155: The definition of an extreme event is an event that is not frequent (in general, above the 99th percentile) and seasonally dependent. The definition used in this sentence is not correct.
Lines 158-164: The authors must follow the WMO guidelines and rewrite the sentence.
Lines 186-188: Does the rainfall threshold refer to the daily accumulation (24h)? Why 15 mm and 5 mm thresholds?
Lines 194-195: Are the mentioned occurrences related to continuous rainfall? How were the values considered? Above 0,1 mm?
Lines 191-204: The authors compare extreme rainfall events with different long-time durations. To compare them, it is essential to consider the same time interval.
Lines 235-237: What references are the authors using to affirm that?
Lines 261-262: The sentence "When the risk of heavy precipitation is influenced by the combination of two 262 variables, for example, the degree of damage caused by heavy precipitation" is wrong. Rainfall is the cause of possible damage. Not otherwise.
Line 286: Using 50 mm for at least six hours is not an objective criterion for characterizing heavy rainfall events. If a precipitation event presents an accumulated volume of 50 mm during 72h, it is not heavy rainfall.
Lines 317-334: The physical nature of the rainfall is different considering the time duration.
Lines 366-367: This manuscript does not discuss the atmospheric conditions that lead to extreme rainfall events. So, the authors cannot extrapolate their results for topics not shown in the manuscript.
Lines 416-417: How do the authors think the proposed methodology can be used to forecast extreme rainfall?
Lines 443—447: It is well known that the higher the rainfall volume and also its duration, the more damages are expected. I don't believe the decision-makers need the results presented in this manuscript in an operational procedure (short-term forecasting).

Reply

Citation: https://doi.org/10.5194/egusphere-2024-2139-RC1

Siyu Liu and Xuguang Dong

Viewed

Total article views: 141 (including HTML, PDF, and XML)

HTML	PDF	XML	Total	BibTeX	EndNote
111	21	9	141	2	3

HTML: 111
PDF: 21
XML: 9
Total: 141
BibTeX: 2
EndNote: 3

Views and downloads (calculated since 09 Sep 2024)

Month	HTML	PDF	XML	Total
Sep 2024	106	18	9	133
Oct 2024	5	3	0	8

Cumulative views and downloads (calculated since 09 Sep 2024)

Month	HTML	PDF	XML	Total
Sep 2024	106	18	9	133
Oct 2024	5	3	0	8

Viewed (geographical distribution)

Total article views: 135 (including HTML, PDF, and XML) Thereof 135 with geography defined and 0 with unknown origin.

Country	#	Views	%

Latest update: 10 Oct 2024

Short summary

The current research and application of the Copula function is mainly limited to two-dimensional; with the advancement of computer technology and the complexity of the research problems, the application of three-dimensional and above Copula function and the accompanying problems of parameter estimation and the choice of the function type need to be further researched. Therefore, the Copula function has great potential in the risk assessment of natural disasters caused by multiple factors.


Total:	0
HTML:	0
PDF:	0
XML:	0