Mapping the potential flood frequency zones in Jordan using Gumbel distribution and GIS

Lubna Al Mahasneh 1; Doaa Abuhamoor 1; Khaldoun Al Sane 2*; Hasan Aldashti 3; Nizar Haddad 4

1, Department of Geographic Information System and Remote Sensing, Directorate of Environment and Climate Change, National Agricultural Research Center (NARC), Amman, Jordan

2, Department of Plant Biodiversity Characterization, Directorate of Plant Biodiversity Research, National Agricultural Research Centre (NARC), Amman, Jordan

3, Department of Meteorology, Directorate General of Civil Aviation, 13001 Kuwait, Kuwait

4, National Agricultural Research Centre (NARC), Amman, Jordan

E-mail:
khaldoun_mk@yahoo.com

Received: 17/09/2024
Acceptance: 29/03/2025
Available Online: 02/04/2025
Published: 01/07/2025

DYSONA – Applied Science

 

Manuscript link
http://dx.doi.org/10.30493/DAS.2025.479057

Abstract

Floods are a major natural hazard in Jordan, causing economic and environmental damage.  Analysis of extreme values of rainfall data using the Gumbel distribution is a useful method in hydrological applications for flood risk management.  This study analyzed 37 years of rainfall data and Gumbel distribution to estimate long-term rainfall and flood frequency.  Return durations and exceedance probability were estimated and mapped for probable flood zones for 25, 50, 75, and 100-year intervals.  Results suggest maximum yearly rainfalls have a 39-year return period with a 2.5% likelihood of recurrence, while medium rainfalls occur annually with a 97% probability.  The results of the spatial analysis indicated that flood frequency, throughout all return periods, was concentrated in the western mountainous regions of the research area, in contrast to the eastern desert areas.  This study determined that extreme rainfall events frequently occur in densely populated regions, posing significant flood risks to the population’s safety and the sustainability of their economic activities.  The findings of this study offer significant insights for decision-makers and planners in mitigating flood risk and facilitating safe, sustainable development in Jordan.

Keywords: Flood frequency, Gumbel distribution, GIS, Jordan

Introduction

In Jordan, floods are a persistent occurrence, resulting in considerable economic and environmental consequences. Flood frequency mapping serves as an effective instrument for evaluating the likelihood of flooding in a certain region. Flash floods are defined by an excessive volume of water occurring within a brief timeframe [1][2]. Future flood risks are anticipated to rise due to a complex interaction among changes in exposures, vulnerabilities, and hazards [3-5]. It is reported that 42% of land regions worldwide would experience heightened flood frequency and reduced return periods [6].

Flood frequency analysis is a technique employed to assess the magnitude of floods during a specified return period [7]. Engineers and hydrologists routinely utilize this technique in the planning, construction, and management of water resource projects and hydraulic structures, including dams, spillways, and flood control systems [8-10]. By applying historical data to probability distributions, future flood occurrences can be anticipated, correlating their intensity with frequency [11]. Historical records assist in estimating hazard recurrence in a certain location [6]. However, these records can be incomplete and restricted to eras when pertinent data, such as seismograph networks, were accessible. Heightened frequency and intensity of precipitation, potentially associated with climate change, are expected to increase flash floods.

Flood probability maps serve as crucial informational resources on flood occurrences. Additionally, flood maps generally have flood zones that categorize continuous and spatially variable threats into distinct flood zones [12-16]. The outer boundary of a zone represents the maximum reach of a flood with a certain probability (e.g., the 1 in 100-year flood), whereas the inner boundary denotes the maximum extent of a flood of lesser magnitude but higher likelihood. The decision to convey information through flood zones may induce a little bias in flood probabilities [14]. Furthermore, ecosystem management is seen as a sustainable strategy for natural resource management that can alleviate the effects of flooding.

Understanding flood probabilities and dangers may assist with flood mitigation planning, hence influencing decision-making [17]. Numerous research studies have investigated the prevalence of flash floods in Jordan. Reports indicated that a flash flood with a 100-year return time in April 1963 represented a considerable risk to Wadi Mousa City [18]. Dry communities like Aqaba, Ma’an, and Wadi Musa-Petra have encountered flash floods of differing intensities: low (5–7 year return intervals), medium (20–25 years), and high (50 years), resulting in considerable damage to life and infrastructure [18]. Another research evaluated flash flood dangers in Jordan, generating a flood hazard severity map that classified 17.6% of the area as “high” hazard, 34.5% as “low,” and 47.9% as “medium,” employing the Rational Method and Integrated Context Analysis (ICA) [19]. In addition to the Rational Method, frequency analysis, unit hydrograph, and rainfall-runoff models are essential tools for comprehending the correlation between flood amplitude and recurrence [11]. This relationship might vary considerably depending on the type of threat, while it remains stable for a specific site. Frequency analysis is an essential statistical method for forecasting extreme hydrological occurrences, such as floods [20]. Likewise, [21] concentrated on calculating flood peaks and proposing mitigating strategies or mapping flood vulnerability in desert areas of Jordan.

Geographic Information System (GIS) technology is increasingly recognized for providing essential geo-scientific data for both regional and site-specific studies [22]. For instance, the relationship between rainfall intensity, duration, and frequency in Jordan’s Mujib basin was explored using the Gumbel method [7]. Records show that arid cities like Aqaba, Ma’an, and Wadi Musa-Petra have experienced flash floods of varying magnitudes: low (5–7-year return periods), medium (20–25 years), and high (50 years), causing significant damage to life and infrastructure [18].

In a related study, hydrological characteristics of severe floods in Aqaba were examined [23]. Low-magnitude floods (6–7-year return periods) and high-magnitude floods (50-year return periods) have been recorded since the 1950s [24]. In 1966, the Central Water Authority noted that the extreme flood event on 11 March 1966 in Ma’an had a 50-year return period. Other notable floods occurred on 8 April 1963 (25-year return period) and 2 February 2006 (19-year return period), with additional floods in 1991, 1993, 1994, 2012, 2013, and 2014. Overall, this study aims to produce flood frequency and probability maps, which are crucial for designing dams, bridges, and flood control structures to mitigate flood disasters and improve stormwater management in Jordan.

Materials and Methods

Study area

Jordan is situated in the Middle East at geographic coordinates 31º 00’N, 36º 00’E (Fig. 1). Covering an area of 89,213 km² with approximately 90% of the country consisted of deserts. Annual precipitation varies from 50 mm in arid areas to 800 mm in the northern hills, where some of the precipitation occurs as snow.

Jordan’s geographic location results in a predominantly hot and arid climate throughout summer, although winters are influenced by Mediterranean low-pressure systems that bring precipitation and snowfall.  The geographic features greatly influence the local climate, augmenting rainfall due to the orographic action of rising and condensing air. 

Mapping the potential flood frequency zones in Jordan using Gumbel distribution and GIS
Figure 1.  Jordan map governorates (Source: https://mapcruzin.com/ Layout prepared by: GIS & RS Department NARC, Jordan). The triangles refer to the spatial distribution of rainfalls gauged stations.

Data collection and processing

Rainfall data

Rainfall data were ordered in descending order over 37 years for historical long-term annual precipitation from 20 gauged stations over the period from 1980 to 2017 (Fig. 1).

Flood frequency analysis approach

The flood frequency study utilized a dataset of long-term annual rainfall spanning 37 years, from 1980 to 2017, collected from 20 locations in Jordan. The frequency of occurrence and probability of long-term yearly rainfall were calculated using the Gumbel method. This distribution is a statistical technique frequently employed to forecast extreme hydrological occurrences, such as floods. It is utilized for flood frequency analysis based on the premise of a high probability of occurrence of maximum yearly rainfall quantities above the mean rainfall over the specified timeframe. The Gumble formula is articulated as follows:

Mapping the potential flood frequency zones in Jordan using Gumbel distribution and GIS

Where (P) is the Frequency (probability of exceedance), (m) is the rank of the item on series, (n) is number of years of record, (T) is the return period of number of years event. The flood frequency mapping was performed using Geographic Information Systems (GIS). ArcGIS 10.5 software was used to produce flood frequency maps.

Spatial interpolation

An interpolation approach was employed to create a thorough representation of the rainfall observations throughout the study area.  The inverse distance weighted (IDW) interpolation technique was utilized with ArcGIS 10.5.  This method evaluates the proximity between observations and their associated attributes to generate a continuous grid of the specified attribute [25].  This approach is extensively acknowledged as the most efficient strategy for achieving enough attribute coverage across expansive study regions.  The annual rainfall data at designated return periods (2, 2-5, 5-10, 10-20, 20-40) were utilized to drive the interpolation procedure.

Results and Discussion

Flood frequency analysis involves correlating flood features with a recurrence interval to determine the probability of occurrence. Subsequently, the probability of exceedance and the frequency of occurrence (return periods) of analogous incidents are calculated based on rainfall data analysis at each station using Gumbel equations. The frequency analysis results indicate that the interpolated maximum annual rainfalls at each station have a return period of 39 years (T), with a very low probability of exceedance for the same event or greater (P) of 2.5%. In contrast, the mean annual rainfalls have a return period of T=1, with a high probability of exceedance of 97% (P). The occurrence frequency is categorized into ranges, and the interpolated maximum rainfall amount over the specified period is presented together with its probability (Fig. 3).

In the 2018/2019 rainy season, Jordan witnessed three significant floods, two of which were deemed the most lethal flash floods in recent decades. The floods resulted in numerous fatalities and injuries, significant infrastructural damage, and economic losses in Amman and other cities. Despite the fact that Jordanian territories experience low to moderate annual precipitation throughout the rainy season, the country occasionally encounters intense rainfall events that typically result in catastrophic flash floods [26].

Reports show that flood index methodology yielded superior flood quantile values for smaller return periods of 2 and 5 years, whereas the direct interpolation method provided more accurate estimations for greater return periods [27]. Moreover, incorporating supplementary atmospheric and land-surface conditions, along with a multi-level model framework that encompasses additional basins in a region, could enhance model performance in arid basins [28].

During the previous few decades, Jordan has had encountered numerous catastrophic flash floods resulting in significant damage to infrastructure, including historical sites, economic losses, and human casualties [29][30]. Climate change scenarios are anticipated to result in an increased frequency of heavy rainfall, hence heightening the risk of flash floods.

The probability distributions frequently employed for flood frequency studies include: Normal, Lognormal, Exponential, Gamma, Pearson Type III, Log-Pearson Type III, generalized extreme value, Pareto, and Gumbel distributions [20][31][32]. Numerous techniques are available for fitting frequency distributions to sample data, including the graphical approach, method of moments, L-moments, maximum likelihood, least squares, and probability weighted moments [33][34]. All these methods utilize sample data to quantify the parameters of statistical distributions [35]. The selection of an appropriate estimation method can result in precise estimation of distribution parameters [36].

The recurrence of flood events is predominantly concentrated in the western regions of the study area.  This observation can be attributed to the elevated topography of these regions relative to the arid desert climate of the central and eastern areas. This observation is particularly significant due to the high population density and intense economic activity in these regions. Consequently, there is an urgent necessity for comprehensive flood risk management measures in these specific areas, which will facilitate sustainable land management and bolster the protection of both population and infrastructure.

The current study utilized the Gumbel distribution to provide critical insights into flood risks in Jordan. The methodology can be further enhanced by incorporating more extensive rainfall datasets and considering additional factors such as land use changes and climate variability.

Mapping the potential flood frequency zones in Jordan using Gumbel distribution and GIS
Figure 3. Maximum annual rainfalls (mm) occurred at return periods of < 2 years with P=51-97% (A), 2 to 5 years with P=21-50% (B), 5-10 years with P=10-20% (C) 10-20 years with P=5-9% (D), and 20-40 years with P=2-4% (E)

Conclusions

This study examined flood frequency and probability in Jordan, highlighting the significance of flood frequency analysis in disaster risk mitigation methods. The application of GIS technology to amalgamate various data sources and generate flood frequency maps facilitates informed hydrological management. Precise flood frequency estimation is essential for floodplain management, crop insurance strategies, public safety, minimizing flood-related expenses, constructing hydraulic structures, and evaluating floodplain hazards, as determined by prior studies. Future research could explore additional remote sensing technologies to enhance flood risk assessments and inform ecosystem management policies. By comparing the current findings with other studies, the effectiveness of the used approach can be validated the need for continuous improvements in flood risk management practices is highlighted.

References

  1. Al-Nammari F, Alzaghal M. Towards local disaster risk reduction in developing countries: Challenges from Jordan. Int. J. Disaster Risk Reduct.  2015;12:34–41. DOI
  2. Gruntfest E, Handmer J. Dealing with Flash Floods: Contemporary Issues and Future Possibilities. In Coping With Flash Floods.  2001;3–10. DOI
  3. Wing OE, Bates PD, Smith AM, Sampson CC, Johnson KA, Fargione J, Morefield P. Estimates of present and future flood risk in the conterminous United States. Environ. Res. Lett.  2018;13(3):034023. DOI
  4. Wu Y, Xue L. Frequency Analysis and Uncertainty Assessment of Annual Maximum Flood Series Using Bayesian MCMC Method. IOP Conf. Ser.: Mater. Sci. Eng.  2018;392:062107. DOI
  5. Diakakis M. flood history analysis and its contribution to flood hazard assessment: the case of Marathonas, Greece. Bull. Geol. Soc. Greece.  2017;43(3):1323. DOI
  6. Hirabayashi Y, Mahendran R, Koirala S, Konoshima L, Yamazaki D, Watanabe S, Kim H, Kanae S. Global flood risk under climate change. Nat. Clim. Change.  2013;3(9):816–21. DOI
  7. Hadadin N. Rainfall Intensity-Duration-Frequency Relationship in the Mujib Basin in Jordan. J. Appl. Sci.  2005;5(10):1777–84. DOI
  8. Erena SH, Worku H. Flood risk analysis: causes and landscape based mitigation strategies in Dire Dawa city, Ethiopia. Geoenvironmental Disasters.  2018;5(1). DOI
  9. Kordrostami S, Alim MA, Karim F, Rahman A. Regional Flood Frequency Analysis Using an Artificial Neural Network Model. Geosciences.  2020;10(4):127. DOI
  10. Mishra V, Cherkauer KA, Shukla S. Assessment of Drought due to Historic Climate Variability and Projected Future Climate Change in the Midwestern United States. J. Hydrometeorol.  2010;11(1):46–68. DOI
  11. Jaiswal P, van Westen CJ. Estimating temporal probability for landslide initiation along transportation routes based on rainfall thresholds. Geomorphology.  2009;112(1–2):96–105. DOI
  12. Alfieri L, Burek P, Dutra E, Krzeminski B, Muraro D, Thielen J, Pappenberger F. GloFAS – global ensemble streamflow forecasting and flood early warning. Hydrol. Earth Syst. Sci.  2013;17(3):1161–75. DOI
  13. Jafarzadegan K, Merwade V. Probabilistic floodplain mapping using HAND-based statistical approach. Geomorphology.  2019;324:48–61. DOI
  14. Ludy J, Kondolf GM. Flood risk perception in lands “protected” by 100-year levees. Nat. Hazards.  2012;61(2):829–42. DOI
  15. Smith BW, Freedy JR. Psychosocial resource loss as a mediator of the effects of flood exposure on psychological distress and physical symptoms. J. Trauma. Stress.  2000;13(2):349–57. DOI
  16. Woznicki SA, Baynes J, Panlasigui S, Mehaffey M, Neale A. Development of a spatially complete floodplain map of the conterminous United States using random forest. Sci. Total Environ.  2019;647:942–53. DOI
  17. Zarekarizi M, Srikrishnan V, Keller K. Neglecting uncertainties biases house-elevation decisions to manage riverine flood risks. Nat. Commun.  2020;11(1). DOI
  18. Al-Weshah RA, El-Khoury F. Flood Analysis and Mitigation for Petra Area in Jordan. J. Water Resour. Plan. Manag.  1999;125(3):170–7. DOI
  19. AlMahasneh L, Abuhamoor D, Al Sane K, Haddad NJ. Assessment and mapping of flash flood hazard severity in Jordan. Int. J. River Basin Manag.  2021;21(2):311–25. DOI
  20. Haberlandt U, Radtke I. Hydrological model calibration for derived flood frequency analysis using stochastic rainfall and probability distributions of peak flows. Hydrol. Earth Syst. Sci.  2014;18(1):353–65. DOI
  21. Farhan Y, Ayed A. Assessment of Flash-Flood Hazard in Arid Watersheds of Jordan. J. Geogr. Inf. Syst.  2017;09(06):717–51. DOI
  22. Elsebaie IH. Developing rainfall intensity–duration–frequency relationship for two regions in Saudi Arabia. J. King Saud Univ. Eng. Sci.  2012;24(2):131–40. DOI
  23. Murphy B. Hydrologic Analysis for a Hyper-Arid Region in the Middle East. Watershed Management Conference 2010: Innovations in Watershed Management under Land Use and Climate Change.  2010. DOI
  24. Baratti E, Montanari A, Castellarin A, Salinas JL, Viglione A, Bezzi A. Estimating the flood frequency distribution at seasonal and annual time scales. Hydrol. Earth Syst. Sci.  2012;16(12):4651–60. DOI
  25. Castiglioni S, Castellarin A, Montanari A. Prediction of low-flow indices in ungauged basins through physiographical space-based interpolation. J. Hydrol.  2009;378(3–4):272–80. DOI
  26. Mazahreh S, Alkharabsheh M, Bsoul M, Hammor DA, Mahasneh LA. Identification and mapping of hotspot areas susceptible to soil erosion in erak al karak area using geoinformatics. Int. J. Res Granthaalayah.  2018;6(6):246–59. DOI
  27. Baidya S, Singh A, Panda SN. Flood frequency analysis. Nat. Hazards.  2020;100(3):1137–58. DOI
  28. Awasthi C, Archfield SA, Ryberg KR, Kiang JE, Sankarasubramanian A. Projecting Flood Frequency Curves Under Near‐Term Climate Change. Water Resour. Res.  2022;58(8). DOI
  29. Human Development Report 2011.  United Nation Development Programme (UNDP). 2011. DOI
  30. National and local disaster risk reduction strategies and plans. 2019- 2022.   United Nation Development Programme (UNDP). 2019. DOI
  31. Gericke O, Du Plessis J. Development of a customised design flood estimation tool to estimate floods in gauged and ungauged catchments. Water SA.  2013;39(1). DOI
  32. Hu L, Nikolopoulos EI, Marra F, Anagnostou EN. Sensitivity of flood frequency analysis to data record, statistical model, and parameter estimation methods: An evaluation over the contiguous United States. J. Flood Risk Manag.  2019;13(1). DOI
  33. Engeland K, Wilson D, Borsányi P, Roald L, Holmqvist E. Use of historical data in flood frequency analysis: a case study for four catchments in Norway. Hydrol. Res.  2017;49(2):466–86. DOI
  34. Noto LV, La Loggia G. Use of L-Moments Approach for Regional Flood Frequency Analysis in Sicily, Italy. Water Resour. Manag.  2008;23(11):2207–29. DOI
  35. Bhatt GD, Sinha K, Deka PK, Kumar A. Flood Hazard and Risk Assessment in Chamoli District, Uttarakhand Using Satellite Remote Sensing and GIS Techniques. Int. J. Innov. Res. Sci. Eng. Technol.  2014;03(08):15348–56. DOI
  36. Bhuyan A, Borah M, Kumar R. Regional Flood Frequency Analysis of North-Bank of the River Brahmaputra by Using LH-Moments. Water Resour. Manag.  2009;24(9):1779–90. DOI

Cite this article:

Al Mahasneh, L., Abuhamoor, D., Al Sane, K., Aldashti, H., Haddad, N. Mapping the potential flood frequency zones in Jordan using Gumbel distribution and GIS. DYSONA – Applied Science, 2025;6(2): 343-350. doi: 10.30493/das.2025.479057