Journal of Geographic Information System, 2011, 3, 225-231
doi:10.4236/jgis.2011.33019 Published Online July 2011 (
Copyright © 2011 SciRes. JGIS
GIS-Based Spatial Mapping of Flash Flood Hazard in
Makkah City, Saudi Arabia
Gomaa M. Dawod1,2, Meraj N. Mirza3, Khalid A. Al-Ghamdi4
1Survey Research Institute, Giza, Egypt
2Umm Al-Qura University, Makkah, Saudi Arabia
3Center of Research Excellence in Hajj and Omrah, Umm Al-Qura
University, Makkah, Saudi Arabia
4Geography Department, Umm Al-Qura University, Makkah, Saudi Arabia
E-mail: dawod_gomaa@yahoo .com,,
Received March 29, 2011; revised May 12, 2011; accepted May 20, 2011
Flash floods occur periodically in Makkah city, Saudi Arabia, due to several factors including its rugged to-
pography and geological structures. Hence, precise assessment of floods becomes a more vital demand in
development planning. A GIS-based methodology has been developed for quantifying and spatially mapping
the flood characteristics. The core of this new approach is integrating several topographic, metrological,
geological, and land use datasets in a GIS environment that utilizes the Curve Number (CN) method of flood
modelling for ungauged arid catchments. Additionally, the computations of flood quantities, such as depth
and volume of runoff, are performed in the attribute tables of GIS layers, in order to assemble all results in
the same environment. The accomplished results show that the runoff depth in Makkah, using a 50-years re-
turn period, range from 128.1 mm to 193.9 mm while the peak discharge vary from 1063 m3/s to 4489 m3/s.
The total flood volume is expected to reach 172.97 million m3 over Makkah metropolitan area. The advan-
tages of the developed methodology include precision, cost-effective, digital outputs, and its ability to be
re-run in other conditions.
Keywords: Flood Assessment, Rainfall-Runoff Model, NRCS, GIS, Saudi Arabia
1. Introduction
Hazards of flash floods are vital in terms of human lives
loss and economical damages. Makkah city, west of
Kingdom of Saudi Arabia (KSA), exhibits two unique
features that increase the hazardous flood consequences:
(1) its topography is very complex; (2) about three mil-
lion Muslims are gathered annually in Makkah to per-
form Hajj over a two-week period in the winter, which is
the main rainfall season in Saudi Arabia. Due to the in-
creasing interest in flood impacts over the last couple of
decades, extensive flood estimation studies have been
carried out in different countries, such as USA [1], Egypt
[2,3], Nigeria [4], South Korea [5], China [6], and Saudi
Arabia [7,8]. This paper aims to develop a GIS approach
for assessing the flash flood hazards for Makkah metro-
politan area, utilizing the most up-to-date and precise
available data sets.
2. Flood Estimation
Flood estimation methods aim to model the rainfall-run-
off relationships, and can be categorized into three
groups according to their complexity. Simple approaches,
such as the rational method and empirical formulas, es-
timate the peak discharge quickly and with little number
of inputs. The Curve Number (CN) is an example of
moderate flood estimation methods. Detailed, or com-
plex, models are able to identify the causes of problems
rather than producing a simple description of overall
conditions [9]. The CN method is quite used in engi-
neering design and flood management projects, particu-
larly in the USA [10-13].
Geographic Information Systems (GIS) and Remote
Sensing (RS) techniques have been utilized as efficient
tools in flood risk assessment [14,15]. For example,
Change et al. [16] applied GIS to study the time-based
relationship between flood hazards and land use changes.
Also, Jasrotia and Singh [17] uutilized the CN method to
study the runoff and soil erosion within a GIS environ-
ment. Moreover, Chen et al. [18] tested a GIS model,
which consists of a storm-runoff model and an inunda-
tion model, to model flood hazards. In addition,
Dongquan et al. [19] developed a GIS batch process to
delineate catchments and compute their geomorphologic
parameters. Furthermore, Guptaa, and Panigrahya [20]
has utilized several data sources and two runoff models
in a GIS platform to investigate the flood characteristics
and variations of large basins in India. Additionally,
Gogoase et al. [21] utilized GIS to develop inundation
maps foe extreme flood events. Moreover, Karmakar et
al. [22] proposed a methodology for six major damage
centers in the Upper Thames River watershed, Canada to
assess the flood risks, i.e. flood probability of occurrence,
vulnerability to flood, and exposures of land use and soil
type to flood.
3. Flash Floods in Makkah City
Makkah city is located in the south-west part of KSA,
about 80 Km east of the Red Sea (Figure 1). It extends
from 39˚35’ E to 40˚02’E, and from 21˚09’ N to 21˚37’
N. The area of the metropolitan region (the study area)
equals 1593 square kilometres. The topography of Mak-
kah is complex in nature, and several mountainous areas
exist inside its metropolitan area. The winter is consid-
ered as the main rainy season in Saudi Arabia. The an-
nual rain over Makkah city, for a period extending from
1966 to 2009, varies from 3.8 mm to 318.5 mm, with an
average of rainfall equals 101.2 mm (Figure 2). Due to
the complexity of Makkah’s topography, flash floods
occur periodically with significant variations in magni-
tude. Mirza and Ahmed [23] have reported that the ex-
treme flood type is repeated with a return period of 46
years, while a second-order flood takes place occasion-
ally with a return period of 33 years, and a low-danger-
ous flood comes about every 13 years. Using the magni-
tude of the annual rain average (which equals 101.2 mm)
as a rain intensity factor might not be optimum in flood
estimation process. The rain intensity of a single extreme
storm may exceed the annual rainfall average for a year.
For example, the 1969 storm records (Figure 3) showed
that the rain intensity reaches 107.5 mm/hour during the
first 10 minutes of that storm. Based on records of a sin-
gle rainfall station, this extreme storm resulted in a run-
off volume of more than 41 million cubic meters in the
central area of Makkah city, with results of severe dam-
ages and human loses [23].
Analyzing the flood series frequency, the return period
or recurrence interval can be computed. That period de-
fines the average number of years during which a flood
of a given magnitude will be equalled or exceeded once.
The Welbull method, among several other formulas,
computes the return period T as [24]:
1Tn m (1)
where n is the number of events, or number of records, m
is the order or rank of the event (flood item) when flood
magnitudes ranked in descending order.
Figure 1. Study area.
Copyright © 2011 SciRes. JGIS
Copyright © 2011 SciRes. JGIS
Figure 2. Annual rains in Makkah city from 1966 to 2009.
Figure 3. Rain intensity of the 1969 storm in Makkah city.
The computed return period of the 1969’s flood has
been estimated to 44 years. That piece of information is
quite helpful in flood assessment studies, as it means that:
1) that flood magnitude is expected to occur by about
2013; and 2) selected return period value for flood man-
agement projects should be equal or greater than 44 years.
The rainfall intensity for a 50-years return period has
been estimated as 200 mm/h [25] and is used in the cur-
rent research study.
4. Data and Methodology
Several datasets have been collected for flood assessment
in Makkah city. The main data set is a Digital Elevation
Model (DEM) over the study area. This DEM was pro-
duced by the by King Abdulaziz City of Sciences and
Technology (KACST) with a spatial resolution equals to
5 meters. A window covering Makkah metropolitan area
(Figure 4) has been provided through the Center of Ex-
cellence in Hajj and Omrah, Umm Al-Qura university.
Mirza et al. [26] confirm that this national DEM is 3
times more accurate than previously published global
DEMs (e.g. ASTER and SRTM) over Makkah area.
From Figure 4, it can be noticed that the heights of
Makkah metropolitan area range from 80 to 982 meters.
The other datasets include digital geological, soil, and
Figure 4. The national 5-m DEM for Makkah city.
land use maps of the study area.
The developed GIS-based flood assessment method-
ology consists of several stages. Figure 5 depicts an
overview flow chart of that scheme. In the first stage, the
Arc GIS software along with the Arc Hydro extension
are used to obtain several shapefiles describing the geo-
morphology of the study area. These shapefiles include:
the main basins and the sub-basins of each main catch-
ment, along with drainage network using Strahler
method (a simple widely-utilized network order method),
and the longest stream path in each catchment. The sec-
ond stage of the developed methodology is based on the
flood assessment method developed by the US Natural
Resources Conservation Service (NRCS), formerly
known as the Soil Conservation Service (SCS). It worth
mentioning that there are several hydrologic methods
used for flood estimation, but the SCS method has been
applied in the current research study. This method, also
known as the Curve Number (CN), makes use of geo-
logical information to assign a unique CN value for each
area, which will be further used to estimate the surface
runoff depth and the peak discharge magnitude. VBA is
used to compute the required flood defining parameters
that consists of [27]:
 
0.2 20.8QP SP S (2)
25.4 100010SCN (3)
qp quAQ
Figure 5. The developed GIS-based flood assessment meth-
here: Q is the depth of direct runoff (mm), P is the
entioning that P was computed through sta-
ydrological parameters include:
depth of precipitation for a specific return period (mm),
S is the maximum potential retention (mm), CN is the
curve number, QT is the volume of runoff (m3), A is the
Area of basin (Km2), Q is the depth of direct runoff (m),
qp is the peak discharge (m3/s), and qu is the unit peak
discharge (m3/s/km2/mm) that can be interpolated from a
specific charts [28] or computed from corresponding
tables [13].
It worth m
tical analysis of rainfall records (e.g. by Log Pearson
III method) to determine the expected rainfall depth for a
specific storm to be occurred again with the same mag-
Other h
0.2279vLtc (6)
0.8 0.5
1900 SL
(7) 1.67 1tcL S
where v is the flow velocity (m/
. Results
he developed GIS-based methodology has been carried
lues found to vary
0.133Sd Tc
s), L is the basin length
(expressed in units of meters), tc is the concentration
time (minutes) that can be estimated using several for-
mulas, one of them (NRCS method) is given in Equation
(7), SL is the average watershed land slope in percentage,
and Sd is the storm duration (hours).
out using the available datasets of Makkah metropolitan
area. Results of the first stage include several maps and
the estimation of many morphometric parameters. It has
been found that there are 6 main catchments in Makkah,
whose areas range from 74.3 to 360.6 square kilometres,
and lengths of their main streams vary from 16.50 to
48.55 kilometres (Figure 6). Table 1 presents statistics
of some morphometric parameters of the six catchments.
The second stage of the developed approach results is
e determination of flood characteristics in Makkah city.
The computations have performed using the depth of
precipitation (P) equals 200 mm for a return period of
50 years. That value is the result of the Log Pearson III
statistical analysis for the available rainfall datasets of
Makkah city. Additionally, the unit peak discharge (qu
in Equation (4)) has been computed through the equation
and tables provided in [13]. Table 2 and Figure 7 pre-
sent the flood estimated parameters.
Moreover, the peak discharge va
m 1063 m3/s (for catchment C2) to 4489 m3/s (for
catchment C5). Additionally, the runoff depths of the six
basins vary from 151.7 mm to 178.8 mm. The flood
volume range between 13.28 m3 (for catchment 3) and
54.69 m3 (for catchment 3), with a total of 172.97 million
m3 over Makkah metropolitan area.
Figure 6. Catchments and their main streams.
Table 1. Statistics of morphometric quantities.
Item C1 C2 C3 C4 C5 C6
Basin Area m2) 73 3 9 62
Basin Premier (mk)134.102.
Basin Length (km) 42.4823.64 16.50 29.70 48.5538.13
134.6 69.13 50.23 89.09 7603
Copyright © 2011 SciRes. JGIS
Copyright © 2011 SciRes. JGIS
Figure 7. Flood spatial variations in Makkah.
able 2. Flood characteristics in Mak
C1 C2 C3 C4 C5 C6
ykah city (for a 50-
ears return period).
Time of concentration
(hours) 5.69 3.76 1.73 2.63 6.724.17
CN 84 84 93 89 84 83
Runoff depth (mm) 1 11 1 11
(million m3)
51.751.7 78.8 66.7 51.7 48.8
Peak discharge (m3/s1554 1063 1307 1234 4489 1514
Storm duration (hour) 0.76 0.50 0.23 0.35 0.89 0.56
Flow velocity (m/s) 28.34 23.86 36.19 42.87 27.44 34.68
38.34 18.55 13.28 18.32 54.6929.79
Volume of runoff
Total = 172.97 million m3
6. Discussion
ticed that catchment C3 has the lowest
the runoff, it has been found that
at is an
irst, it can be noF
concentration time (1.98 hours), which is the time for
runoff to travel from the most distant point to the outlet
point. The lower time of concentration, the more haz-
ardous is the runoff. Additionally, catchment C3 is ex-
pected to have minimum storm duration (0.26 hours)
while catchment C1 has the maximum value (0.80 hours).
Concerning the flow velocity of the six basins, it has
been found that the values range from 22.44 m/s (for
catchment C2) to 73.37 m/s (for catchment C5).
Keeping in mind that the lower concentration time, the
catchment C3 has the lower concentration time Th
more hazardous is
cepted result, since catchment C3 has the highest relief
ratio (40.43 m/km) even though it does not have the
highest flow velocity. Additionally, catchment C3 is ex-
pected to have minimum storm duration (0.26 hours)
while catchment C1 has the maximum value (0.80 hours).
It is worth mentioning that the storm duration is a func-
tion of the concentration time as seen in Equation (8).
A coloration analysis was performed between the main
flood factors. The correlation matrix is presented in Ta-
ble 3. Concerning the flow velocity of the six basins, it
has been found that the minimum velocity belongs to
catchment C2 while the maximum velocity is for catch-
ment C5. Equation (5) shows that the flow velocity is
directly proportional to the catchment stream length, and
is inversely proportional to the concentration time. The
same conclusion is drawn from Table 3 where the peak
discharge, total flood volume, basin area, and basin
stream length have positive strong correlations with the
flow velocity (0.97, 0.76, 0.74, and 0.61 respectively),
while the concentration time has a negative moderate
correlation (–0.48).
Second, catchment C3 also has the highest CN value
of 93. The higher CN values, in the study area, is attrib-
uted to two factors: 1) the residential area of Makkah city,
paved streets, was assigned CN of 98 that reflects the
low permeability of rains; 2) a great portion of Makkah’s
geology consists of Pre-Cambrian igneous and meta-
morphic rocks, that get relatively high CN according to
Table 3. Correlation between main flood characteristics.
CN tc A L Q QT v
CN 1
tc –0.65 1
A – 0
– 0 0.
– 0. 0.0
0.0.4 0.
0 0. 0
L 0.72.39941
Q –0.30 0.338571
QT –0.62 0.18 9987 9871
v –0.14 –0.48 0.74 .6197.761
whe CN bise oentn, A the
bas brehunh, roff
voland v is the
the SRC classification.
ast basin area. That leads to this catchment produces
ge. As expected, catchment C5
to the
(e.g. the rational method)
distinguished by two items (a complex topography and
million Muslims to perform
ajj over a specific short time period annually) that
o acknowledge the financial
upport offered by the Center of Research Excellence in
Umm Al-Qura university,
audi Arabia.
. Geological Survey), “Watershed Models for
Decision Support for Inflows to Potholes Reservoir,”
cientific Investigations Report 2009-5081,
re: is the curve numer, tc the timf concratiois
in area,
L is theasin st
runoff ve
am lengt
, Q is roff deptQT is theun
Moreover, it can be noticed that four catchments (C1,
C2, C5, and C6) have similar CN values, but C2 has the
the minimum peak dischar
oduces the highest peak discharge because of its big
basin area. Table 3 concludes the same result since the
basin area and basin stream length has strong correla-
tions (0.85 and 0.70) with the peak discharge.
Additionally, it can be concluded that the highest run-
off depth belong to catchment C3, which has the highest
CN value. Regarding the flood volume, it can be seen
that catchment C5 produces the biggest value due
ct that it has the biggest area (360.6 km2). The correla-
tion between flood volume and basin area is strong
(0.9987) as seen in Table 3.
The GIS-based CN flood estimation methodology has
several advantages. First, it incorporates many input
datasets including metrological, geological, soil, land use,
and DEM. Other approaches
ploy less input items, while some national regression
models used in Saudi Arabia depend only on the basin
area to estimate expected flood. Dawod et al. [29] con-
cluded that the CN method is more precise than some
other flood estimation methods over Makkah region.
Moreover, this paper presents a GIS-based implementa-
tion of the CN methodology where computations have
been performed in the attribute tables within the GIS
environment (using VBA). So, the flood estimation
process is efficient, faster, and can be easily performed
for several scenarios and even for several regions in
Saudi Arabia once the required input layers are available.
7. Conclusions
Makkah metropolitan area, south west of Saudi Arabia,
the gathering of about three
greatly raises awareness of flood hazards impacts. This
research develops a GIS-based approach for mapping
and quantifying flood assessment measures. The devel-
oped methodology is based on integrating several data-
sets in a GIS environment utilizing the SRC CN flood
modelling method. Results show that the main factors
affect the total flood volumes, in Makkah metropolitan
area, are the catchment area, the basin stream length, and
the peak discharge. Additionally, it has been concluded
that the higher CN value (for low permeability soil, geo-
logical, and land use features), the higher runoff and
flood hazards. Merits of that methodology include preci-
sion, cost-effective, digital outputs, and its ability to be
re-run for other scenario (e.g., other design return period).
Hence, it is recommended that the attained results be
utilized in governmental planning in Makkah city, and
that approach should be applied to all other cities in
Saudi Arabia.
8. Acknowledgements
The authors would like t
Hajj and Omrah (Hajjcore),
9. References
[1] USGS (U.S
Washington S
M. Azab, “Assessment and Management of Natural Haz-
ards and Disasters along Qena-Safaga Road, Central
Eastern Desert, Egypt,” Egyptian Journal of Remote
Sensing and Space Science, Vol. 12, 2009, pp. 55-70.
[3] M. El-Bastawesy, K. White and A. Nasr, “Integration of
Remote Sensing and GIS for Modelling Flash Floods in
Wadi Hudain Catchment, Egypt,” Hydrological Process,
Vol. 23, No. 9, 2009, pp. 1359-1368.
[4] S. Yahaya, N. Ahmad and R. Abdalla, “Multicriteria
Analysis for Flood Vulnerable Areas In Hadejia-
Jama’are River Basin, Nigeria,” European Journal of
Scientific Research, Vol. 42, No. 1, 2010, pp. 71-83.
[5] N. Kim, J.W. Lee, J., Lee and J. E. Lee, “SWAT Appli-
cation to Estimate Design Runoff Curve Number for
South Korean Conditions,” Hydrological Process, Vol.
24, No. 15, 2010, pp. 2156-2170.
[6] W. Liang, C. Yongli, C. Hongquan, D. Daler, Z. Jingmin,
and Y. Juan, “Flood Disaster in Taihu Basin, China:
Causal Chain and Policy Option Analyses,” Environ-
mental Earth Science, Vol. 63, No. 5, 2010, pp. 1119-
Copyright © 2011 SciRes. JGIS
Copyright © 2011 SciRes. JGIS
1124. doi:10.1007/s12665-010-0786-x
[7] M. Al Saud, “Assessment of Flood Hazard of Jeddah
Area 2009, Saudi Arabia,” Journal of Water Resource
and Protection, Vol. 2, No. 9, 2010, pp. 839-847.
[8] E. Fred and B. Mostafa, “Flood Risk Modeling for Holy
Sites in Makkah,” Proceedings of the Symposium on Na-
tional GIS application in Saudi Arabia, Al-Kober Cit
2008, pp. 1-12.
y,” Environmental Earth Science, Vol. , No. ,
te Sensing and Space Sciences,
[9] E. Mihalik, “Watershed Characterization and Runoff
Modelling of the Chapel Branch Creek, Orangeburg
county, South Carolina,” MS.C Thesis, College of
Charleston, Charleston, 2008.
[10] IDoT (Iowa State Department of Transportation), “Iowa
Storm Water Management Manual,” 2009.
[11] US ACE (US Army Corps of Engineers), “Physical
Stream Assessment,” 2004.
[12] NH DoT (New Hampshire Department of Transportation),
“Drainage Design Manual for Highways,” 1998.
[13] US DoT (US Department of Transportation), “Highway
Hydrology Manual,” Publication No. FHWA-02-001,
[14] A. Youssef, B. Pradhan and A. Hassan. “Flash Flood
Risk Estimation along the St. Katherine Road, Southern
Sinai, Egypt Using GIS Based Morphometry and Satellite
April 27, 2010, pp. 1-13.
[15] A. Saleh and S. Al-Hatrushi. “Torrential Flood Hazards
Assessment, Management, and Mitigation, in Wadi Aday,
Muscat area, Sultanate of Oman, a GIS & RS Approach,”
Egyptian Journal of Remo
12, No. 1, 2010, pp. 71-86.
[16] H. Chang, J. Franczyk and C. Kim, “What is Responsible
for Increasing Flood Risks? The Case of Gangwon Prov-
ince, Korea,” Natural Hazards, Vol. 48, No. 3, 2009, pp.
339-354. doi:10.1007/s1106
[17] A. Jasrotia and R. Singh, “Modeling Runoff and Soil
Erosion in a Catchment Area, Using the GIS, in the Hi-
malayan Region, India,” Environmental Geology, Vol. 51,
No. 1, 2006, pp. 29-37. doi:10.1007/s00254
[18] J. Chen, A. Hill and L. Urbano, “A GIS-Based Model for
Urban Flood Inundation,” Journal of Hydrology, Vol.
373, No. 1-2, 2010, pp. 184-192.
[19] Z. Dongquan, C. Jining, W. Haozh
Shangbing and S. Zheng, “GIS-B
eng, T. Qingyuan, C.
ased Urban Rainfall-
Runoff Modeling Using an Automatic Catchment- Dis-
cretization Approach: A Case Study in Macau,” Envi-
ronmental Earth Sciences, Vol. 59, No. 2, 2009, pp. 465-
472. doi:10.1007/s12665-009-0045-1
[20] P. Guptaa and S. Panigrahya, “Geo-Spatial Modeling of
Runoff of Large Land Mass: Analysis, Approach and
nts at the Mouth of the Danube
for Risk-Vulnerability Assessment to
rabic),” Geographic Mes-
New Age International Ltd., New Delhi, 2006.
ethodologies in Makkah Met-
Results for Major River Basins of India,” Proceedings of
the International Archives of the Photogrammetry, Re-
mote Sensing and Spatial Information Sciences, Vol. 37,
Part B2, Beijing, 2008.
[21] D. Gogoase, I. Armaş and C. Ionescu, “Inundation Maps
for Extreme Flood Eve
River,” International Journal of Geosciences, Vol. 2, No.
1, 2011, pp. 68-74.
[22] S. Karmakar, S. Simonovic, A. Peck and J. Black, “An
Information System
Flood,” Journal of Geographic Information System, Vol.
2, No. 2, 2010, pp. 129-146.
[23] M. Mirza and B. Ahmed, “Winter Weather and Climate
Conditions in Makkah (in A
sages, Kuwait Geographic Association, Kuwait, Vol. 253,
[24] H. Raghunath, “Hydrology: Principles, Analysis, and
[25] KSA MoPWH (Saudi Ministry of Public Works and
Housing), “Storm Water Drainage System for Muzdalifah
and Arafat Areas,” Unpublished Technical Report, 2001.
[26] M. Mirza, G. Dawod and K. Al-Ghamdi, “Assessment of
Global and National Digital Elevation Models For Geo-
detic and Geomorphologic Applications in Makkah Met-
ropolitan Area, Saudi Arabia,” The XXV IUGG General
Assembly, Melbourne, 28 June - 7 July, 2011.
[27] Z. Sen, “Wadi Hydrology,” CRC Press, New York, 2008.
[28] US NRCS (US National Resources Conservation Ser-
vices), “Urban Hydrology for Small Watersheds,” Tech-
nical Manual TR55, 1986.
[29] G. Dawod, M. Mirza and K. Al-Ghamdi, “Assessment of
Several Flood Estimation M
ropolitan Area, Saudi Arabia,” Under Review, Arabian
Journal of Geosciences, 2011.