**Abstract**

Tropical cyclones [TCs] are a common natural hazard that have significantly impacted Oman. Over the period 1881–2019, 41 TC systems made landfall in Oman, each associated with extreme winds, storm surges and significant flash floods, often resulting in loss of life and substantial damage to infrastructure. TCs affect Omani coastal areas from Muscat in the north to Salalah in the south. However, developing a better understanding of the high-risk regions is needed, and is of particular interest in disaster risk reduction institutions in Oman. This study aims to find and map TC tracks and their spatio-temporal distribution to landfall in Oman to identify the high-risk areas. The analysis uses Kernel Density Estimation [KDE] and Linear Direction Mean [LDM] methods to better identify the spatio-temporal distribution of TC tracks and their landfall in Oman. The study reveals clear seasonal and monthly patterns. This knowledge will help to improve disaster planning for the high-risk areas.

**Keywords:** Oman, Arabian Sea, tropical cyclone, storm track analysis, natural hazard risk assessment, hazard mapping

### **1. Introduction**

Understanding and forecasting the consequences of climate change is critical to support the work of planners and decisions makers. Climate change has touched many aspects of our environment and life; for example, 43% of all natural disasters are related to flooding, which has impacted 56% of all people around the world [1]. Rising global temperature drives more frequent and extreme rainfall events [2], and through thermal expansion, sea level rise. It has been reported that the temperature of the Indian Ocean is increasing faster than other oceans [3]. These factors increase flood risk for coastal zones, particularly those in low lying areas, such as Oman.

These increased risks are expected to persist. For example, in Japan, general circulation models used to forecast the frequency of future storms and cyclones indicate potential losses of about 10 billion USD per year from 100 year return period rainfall events [4]. Similar findings from more frequent and extreme events are projected for central India [5]. Major storms also heavily impact important coastal habitat and ecosystems, particularly coral reefs [6, 7] that are also under climate change pressure from ocean warming and acidification. Recent studies

provide a projection of tropical cyclones (TCs) for 2081–2100 at global and regional scales, and conclude that with a scenario of 2 degrees of global warming, there will be a global reduction in the number of TCs but an increase in TC intensity (more cyclonic events of category 4 and 5), and an increase of tropical cyclonic rainfall amount by 5–20% [8].

IMD e-Atlas for storm and depression tracks over the North Indian Ocean for 1891– 2019 [27]. The data used represent all known major tropical systems in the Arabian Sea, 1881–2019, for which observations are available. The tracks show tropical system intensity [category] with data organized based on the temporal [seasonal and monthly] distribution of cyclone tracks and their point of origin. Tracks that made landfall in Oman were extracted from the total tracks for dedicated analysis. Two methods were used in the spatio-temporal analysis. First, kernel density estimation [KDE] was used to calculate the TCs density distribution. The standard deviations of the KDE showed the high-density area of tracks and were calculated seasonally and monthly for both total Arabian Sea cyclones, and those making

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard…*

Next, the Linear Direction Mean [LDM] is used to identify the linear trend of tracks [the mean orientation and mean angular direction of cyclone tracks], also by month and season. Adding the temporal dimension is useful as it can reveal seasonal variability in tracks. The LDM statistic is "the angle of a line representing the mean direction or orientation of all the lines in the dataset" [28]. The orientation mean considers only the tracks' movements, whilst the linear direction mean ½*θR*] additionally considers the from/to [e.g. east to west] direction of travel. Esri ArcGIS was used to calculate the LDM and circular statistics of compass angle and circular variance [see below] with tracks defined by each cyclone's origin and endpoint coordinates plotted in a space graph with origin 0, 0 [28]. The statistics calculated

*Y* ¼

*X* ¼

*OR* ¼

sum of the sines of track angles. The mean angular direction *θ<sup>R</sup>* is then:

*θ<sup>R</sup>* ¼ arctan

the directional mean, and is analogous to the standard deviation. The circular

subsequently distance, where 1 DD = 111.3 km].

P*<sup>n</sup>*

P*<sup>n</sup>*

*<sup>i</sup>*¼<sup>1</sup> sin *<sup>θ</sup><sup>i</sup> n*

*<sup>i</sup>*¼<sup>1</sup> cos *<sup>θ</sup><sup>i</sup> n*

ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi *<sup>X</sup>*<sup>2</sup> <sup>þ</sup> *<sup>Y</sup>*<sup>2</sup> <sup>p</sup>

Where X and Y are the rectangular coordinates of the mean point, and n is the

Where OR is the length of the resultant vector [in decimal degree [DD] and

cos *<sup>θ</sup>* <sup>¼</sup> *<sup>X</sup>*

sin *<sup>θ</sup>* <sup>¼</sup> *<sup>Y</sup>*

Where *cos θ* is the sum of the cosines of the angles of the trackand *sin θ* is the

The *θ<sup>R</sup>* of the resultant vector ½ � *R* is the mean directional counter clockwise from due east [with a value up to 180°]. The resultant vector ½ � *R* shifted into the correct quadrant is the compass angle clockwise from due north [with value 0–360°]. The circular variance [*s*] shows how much storm track directions deviate from

sin *θ* cos *θ* (1)

(2)

(3)

*OR* (4)

*OR* (5)

� � (6)

landfall in Oman.

*DOI: http://dx.doi.org/10.5772/intechopen.96961*

are as follows:

number of tracks.

**147**

These more intense events are often very damaging. For example, the Gulf Coast of the United States was hit by Hurricane Katrina in 2005, resulting in 1300 fatalities, and infrastructure and urban damage costs of \$75 billion. In 2003, Hurricane Isabel struck North Carolina and New Jersey resulting in 40 deaths and \$3.6 billion in damage [9]. The damage risk is exacerbated by urbanization, including in the Gulf countries, as flood hazards are elevated by modification of otherwise natural land surfaces, whilst more people and assets are also exposed to the elevated flood risk [10].

These pressures apply in Oman, where TC's represent a significant risk to people and infrastructure [11]. TCs and storms coming from the Arabian Sea are common in Oman [12, 13]. They are associated with intense rainfall, flash flooding, and can generate tremendous infrastructural, socio-economic and environmental losses [14–16]. For example, TC Gonu, the first category 5 'super cyclone' recorded in the Arabian Sea in a century, hit northern Oman and Iran in 2007, with 78 fatalities, and an estimated \$4.6 billion in damage [12]. Gonu is considered Oman's worst natural disaster but smaller events can be very damaging too; for example a 2002 cyclonic storm (ARB 01) caused about \$50 million in damage [17, 18]. The most intense recent events to impact Oman were in the 2015 North Indian Ocean season, comprising the cyclonic storm Ashobaa and the extreme cyclonic storms Chapala and Megh, both of which resulted in fatalities. Since then southern Oman has experienced flash flooding due to TCs Mekuno and Luban in 2018, with further intense storm events in 2020 [19]. A higher frequency of extreme rainfall events is now considered the new norm in Oman [20].

In the Arabian Sea region, the majority of TCs form near the Laccadive Islands [11° N, 73° E] in two seasons: the pre-monsoon and the post-monsoon [21]. The pre-monsoon season runs from the end of April to June when the south-west wind rises, and the sea surface becomes very warm. The post-monsoon season runs from September to December, when the south-west wind declines and a northeast wind develop over the Arabian Sea [22].

Recent studies report an increase in the frequency of extremely severe cyclonic storms in the pre-monsoon period [23] an increase attributed to elevated anthropogenic black carbon and sulphate emissions [24]. These anthropogenic emissions lead to a weakening of the climatological vertical wind shear, causing an increase in TC intensity in the Arabian Sea [24].

Most of these more intense TCs make landfall [24] and so pose a growing risk. Understanding cyclone frequency and direction is thus essential in identifying areas at high risk in Oman and supporting disaster management. Therefore, this study analyses TCs in the Arabian Sea to better understand their frequency and direction. The study presents a spatio-temporal analysis of cyclone tracks in the Arabian Sea region, drawing on observations from 1881–2019. The research aims to identify high-frequency seasons, the cyclone direction in each season, and the linear direction trend. The results are intended to identify the more exposed areas around the Arabian Sea, particularly in Oman, and to support disaster risk appraisal and management.

### **2. Data and methods**

A spatial [GIS] database of TCs was created based on tracks obtained from the Indian Metrological Department Atlas for the period 1881–1999 [25, 26] and the

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard… DOI: http://dx.doi.org/10.5772/intechopen.96961*

IMD e-Atlas for storm and depression tracks over the North Indian Ocean for 1891– 2019 [27]. The data used represent all known major tropical systems in the Arabian Sea, 1881–2019, for which observations are available. The tracks show tropical system intensity [category] with data organized based on the temporal [seasonal and monthly] distribution of cyclone tracks and their point of origin. Tracks that made landfall in Oman were extracted from the total tracks for dedicated analysis.

Two methods were used in the spatio-temporal analysis. First, kernel density estimation [KDE] was used to calculate the TCs density distribution. The standard deviations of the KDE showed the high-density area of tracks and were calculated seasonally and monthly for both total Arabian Sea cyclones, and those making landfall in Oman.

Next, the Linear Direction Mean [LDM] is used to identify the linear trend of tracks [the mean orientation and mean angular direction of cyclone tracks], also by month and season. Adding the temporal dimension is useful as it can reveal seasonal variability in tracks. The LDM statistic is "the angle of a line representing the mean direction or orientation of all the lines in the dataset" [28]. The orientation mean considers only the tracks' movements, whilst the linear direction mean ½*θR*] additionally considers the from/to [e.g. east to west] direction of travel. Esri ArcGIS was used to calculate the LDM and circular statistics of compass angle and circular variance [see below] with tracks defined by each cyclone's origin and endpoint coordinates plotted in a space graph with origin 0, 0 [28]. The statistics calculated are as follows:

$$Y = \frac{\sum\_{i=1}^{n} \sin \theta i}{n} \tag{1}$$

$$X = \frac{\sum\_{i=1}^{n} \cos \theta i}{n} \tag{2}$$

Where X and Y are the rectangular coordinates of the mean point, and n is the number of tracks.

$$OR = \sqrt{X^2 + Y^2} \tag{3}$$

Where OR is the length of the resultant vector [in decimal degree [DD] and subsequently distance, where 1 DD = 111.3 km].

$$\cos \overline{\theta} = \frac{X}{OR} \tag{4}$$

$$
\sin \overline{\theta} = \frac{Y}{OR} \tag{5}
$$

Where *cos θ* is the sum of the cosines of the angles of the trackand *sin θ* is the sum of the sines of track angles. The mean angular direction *θ<sup>R</sup>* is then:

$$\theta\_{\mathsf{R}} = \arctan\left(\frac{\sin \bar{\theta}}{\cos \bar{\theta}}\right) \tag{6}$$

The *θ<sup>R</sup>* of the resultant vector ½ � *R* is the mean directional counter clockwise from due east [with a value up to 180°]. The resultant vector ½ � *R* shifted into the correct quadrant is the compass angle clockwise from due north [with value 0–360°].

The circular variance [*s*] shows how much storm track directions deviate from the directional mean, and is analogous to the standard deviation. The circular

provide a projection of tropical cyclones (TCs) for 2081–2100 at global and regional scales, and conclude that with a scenario of 2 degrees of global warming, there will be a global reduction in the number of TCs but an increase in TC intensity (more cyclonic events of category 4 and 5), and an increase of tropical cyclonic rainfall

These more intense events are often very damaging. For example, the Gulf Coast of the United States was hit by Hurricane Katrina in 2005, resulting in 1300 fatalities, and infrastructure and urban damage costs of \$75 billion. In 2003, Hurricane Isabel struck North Carolina and New Jersey resulting in 40 deaths and \$3.6 billion in damage [9]. The damage risk is exacerbated by urbanization, including in the Gulf countries, as flood hazards are elevated by modification of otherwise natural land surfaces, whilst more people and assets are also exposed to the elevated flood risk [10]. These pressures apply in Oman, where TC's represent a significant risk to people and infrastructure [11]. TCs and storms coming from the Arabian Sea are common in Oman [12, 13]. They are associated with intense rainfall, flash flooding, and can generate tremendous infrastructural, socio-economic and environmental losses [14–16]. For example, TC Gonu, the first category 5 'super cyclone' recorded in the Arabian Sea in a century, hit northern Oman and Iran in 2007, with 78 fatalities, and an estimated \$4.6 billion in damage [12]. Gonu is considered Oman's worst natural disaster but smaller events can be very damaging too; for example a 2002 cyclonic storm (ARB 01) caused about \$50 million in damage [17, 18]. The most intense recent events to impact Oman were in the 2015 North Indian Ocean season, comprising the cyclonic storm Ashobaa and the extreme cyclonic storms Chapala and Megh, both of which resulted in fatalities. Since then southern Oman has experienced flash flooding due to TCs Mekuno and Luban in 2018, with further intense storm events in 2020 [19]. A higher frequency of extreme rainfall events is

In the Arabian Sea region, the majority of TCs form near the Laccadive Islands [11° N, 73° E] in two seasons: the pre-monsoon and the post-monsoon [21]. The pre-monsoon season runs from the end of April to June when the south-west wind rises, and the sea surface becomes very warm. The post-monsoon season runs from September to December, when the south-west wind declines and a northeast wind

Recent studies report an increase in the frequency of extremely severe cyclonic storms in the pre-monsoon period [23] an increase attributed to elevated anthropogenic black carbon and sulphate emissions [24]. These anthropogenic emissions lead to a weakening of the climatological vertical wind shear, causing an increase in

Most of these more intense TCs make landfall [24] and so pose a growing risk. Understanding cyclone frequency and direction is thus essential in identifying areas at high risk in Oman and supporting disaster management. Therefore, this study analyses TCs in the Arabian Sea to better understand their frequency and direction. The study presents a spatio-temporal analysis of cyclone tracks in the Arabian Sea region, drawing on observations from 1881–2019. The research aims to identify high-frequency seasons, the cyclone direction in each season, and the linear direction trend. The results are intended to identify the more exposed areas around the Arabian Sea, particularly in Oman, and to support disaster risk appraisal and management.

A spatial [GIS] database of TCs was created based on tracks obtained from the Indian Metrological Department Atlas for the period 1881–1999 [25, 26] and the

amount by 5–20% [8].

*Agrometeorology*

now considered the new norm in Oman [20].

develop over the Arabian Sea [22].

TC intensity in the Arabian Sea [24].

**2. Data and methods**

**146**

variance is calculated from the length of the resultant vector [*OR*], with the result then subtracted from 1 as [28]:

$$s = 1 - \frac{OR}{n} \tag{7}$$

**3.2 Track analysis of Arabian Sea storms making landfall in Oman**

*Monthly distribution of tropical systems in the Arabian Sea, 1881–2019.*

*DOI: http://dx.doi.org/10.5772/intechopen.96961*

The tracks analysis (**Figure 1**) reveals a distinct difference in pre-and postmonsoon cyclones. **Figure 1**(1) shows the distribution of cyclone origin point and track in the pre-monsoon by month. All pre-monsoon cyclone origins were in the Arabian Sea, to the south-east in May, moving slightly to the northeast in June, then to the north in July. The tracks vary in each month direction, but a clear pattern of track movement is to the south-west Arabian Sea in May and to the north-west in June and July, although there are numerous cases in June when the track curves

**Years Jan Feb Mar Apr May June July Aug Sep Oct Nov Dec Total %** –1894 0 0 0 1 2 5 0 0 0 1 3 0 12 5 –1909 0 0 0 1 4 5 2 0 0 5 1 1 19 8 –1924 0 0 0 1 3 2 0 0 0 2 1 0 9 6 –1939 1 0 0 0 5 7 0 0 2 5 4 0 24 9 –1954 0 0 0 4 2 6 0 0 1 4 5 1 23 10 –1969 0 0 0 0 6 5 3 2 1 5 5 2 29 13 –1984 0 0 0 1 6 9 4 0 2 12 10 2 46 21 –1999 0 0 0 0 3 6 0 0 1 6 5 3 24 10 –2014 0 0 0 0 5 9 0 0 4 6 7 1 32 12 –2019 0 0 1 0 2 4 0 0 1 5 1 4 18 6 Total 1 0 1 8 38 58 9 2 12 51 42 14 236 100 % 0% 0% 0% 3% 38% 25% 4% 1% 5% 22% 18% 6% 100

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard…*

**Figure 1**(2) shows the distribution of origin points and tracks in the postmonsoon. Origin points are distributed over a larger area than in the pre-monsoon. Most post-monsoon storms originate in the northeast Arabian Sea in September and move gradually to the south-east and the south Arabian Sea in October–December. However, several storms formed in the Bay of Bengal and track west over India before arriving in the Arabian Sea. The tracks analysis shows that in the postmonsoon, cyclones usually track to the west of the Arabian Sea in September, then gradually to the south-west toward the Gulf of Aden and the Horn of Africa in October–December. However, some November cyclones recurved to India in the

The origins and tracks for the sub-set of systems that made landfall in Oman also vary by season and month of formation. **Figure 1**(1) shows pre-monsoon origins and tracks, revealing a distinct difference in tracks within this period. In May, systems frequently travel to the coast between Masirah Island (central Oman) and Salalah in south-west Oman. In June the track direction moves to central to northeast Oman, such that tropical systems arrive at the coast between Masirah Island and Ras Al Had, the easterly most point of Oman at the mouth of the Oman Sea. Note, however, that the historical record indicates that storms occasionally deviate from this general pattern. A strong storm of May 1898 crossed Oman from Ras

*3.2.1 Tracks classification*

*Source: [IMD archive].*

**Table 2.**

northeast toward India.

east of the Arabian Sea.

**149**

The circular variance, *s*, has a value between 0 and 1. If all lines pointed in the same direction, the *OR* would equal the number of lines n [*OR*/n = 1], and the circular variance [*s*] = 0. If lines pointed in the opposite direction, then *OR* = 0 and *s* =1 [28].

The KDE and LDM analyses thus show the location-frequency [density] of the Arabian Sea tropical cyclones [1881–2019] and their mean tracks, and how these factors vary over time, by cyclone intensity, and in particular for those cyclones which pose the most significant risk to Oman, those making landfall.

#### **3. Results**

#### **3.1 Arabian Sea tropical cyclones frequency**

**Table 1** presents summary statistics for TCs that formed in the Arabian Sea, 1881–2019. In total, 236 systems formed in the Arabian Sea; of which 134 made landfall, and 102 died in the Arabian Sea or the Gulf of Aden. India has the highest frequency of Arabian Sea cyclones making landfall, with 26.7% of the total, about half of all those making landfall. Overall 47 systems made landfall in Oman [19.9% of the total], and another 16 entered Omani coastal waters but died at sea [between 60–64° E].

**Table 2** shows the monthly distribution of Arabian Sea cyclones. The data is grouped into temporal bands that indicate two to four cyclones occur every 15 years, except for 1955–1984 when twice this rate occurred. There is a high frequency of cyclone formation in May–June and October–November. Half of all landfall events occurred in the pre-monsoon [13 in May, nine in June], and 30% in the post-monsoon [five in September, six in October, plus one in November], a pattern consistent with the shorter record of Membery [29]. Thus the formation of TCs in the Arabian Sea occurs in two distinct seasons; pre-and post-monsoon.


#### **Table 1.**

*Distribution of tropical systems in the Arabian Sea, 1881–2019.*

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard… DOI: http://dx.doi.org/10.5772/intechopen.96961*


#### **Table 2.**

variance is calculated from the length of the resultant vector [*OR*], with the result

*<sup>s</sup>* <sup>¼</sup> <sup>1</sup> � *OR n*

The circular variance, *s*, has a value between 0 and 1. If all lines pointed in the same direction, the *OR* would equal the number of lines n [*OR*/n = 1], and the circular variance [*s*] = 0. If lines pointed in the opposite direction, then *OR* = 0

The KDE and LDM analyses thus show the location-frequency [density] of the Arabian Sea tropical cyclones [1881–2019] and their mean tracks, and how these factors vary over time, by cyclone intensity, and in particular for those cyclones

**Table 1** presents summary statistics for TCs that formed in the Arabian Sea, 1881–2019. In total, 236 systems formed in the Arabian Sea; of which 134 made landfall, and 102 died in the Arabian Sea or the Gulf of Aden. India has the highest frequency of Arabian Sea cyclones making landfall, with 26.7% of the total, about half of all those making landfall. Overall 47 systems made landfall in Oman [19.9% of the total], and another 16 entered Omani coastal waters but died at sea [between

**Table 2** shows the monthly distribution of Arabian Sea cyclones. The data is grouped into temporal bands that indicate two to four cyclones occur every 15 years, except for 1955–1984 when twice this rate occurred. There is a high frequency of cyclone formation in May–June and October–November. Half of all landfall events occurred in the pre-monsoon [13 in May, nine in June], and 30% in the post-monsoon [five in September, six in October, plus one in November], a pattern consistent with the shorter record of Membery [29]. Thus the formation of TCs in the Arabian Sea occurs in two distinct seasons; pre-and post-monsoon.

**Landfall country Frequency %** India 63 26.7 Oman 47 19.9 Pakistan 10 4.2 Somalia 7 3.0 Yemen [Socotra islands] 7 3.0 Terminates at Sea 102 43.2 Arabian Sea 96 40.7 Gulf of Aden 6 2.5 Total 236 100

which pose the most significant risk to Oman, those making landfall.

**3.1 Arabian Sea tropical cyclones frequency**

(7)

then subtracted from 1 as [28]:

and *s* =1 [28].

*Agrometeorology*

**3. Results**

60–64° E].

*Source: [IMD archive].*

*Distribution of tropical systems in the Arabian Sea, 1881–2019.*

**Table 1.**

**148**

*Monthly distribution of tropical systems in the Arabian Sea, 1881–2019.*

#### **3.2 Track analysis of Arabian Sea storms making landfall in Oman**

#### *3.2.1 Tracks classification*

The tracks analysis (**Figure 1**) reveals a distinct difference in pre-and postmonsoon cyclones. **Figure 1**(1) shows the distribution of cyclone origin point and track in the pre-monsoon by month. All pre-monsoon cyclone origins were in the Arabian Sea, to the south-east in May, moving slightly to the northeast in June, then to the north in July. The tracks vary in each month direction, but a clear pattern of track movement is to the south-west Arabian Sea in May and to the north-west in June and July, although there are numerous cases in June when the track curves northeast toward India.

**Figure 1**(2) shows the distribution of origin points and tracks in the postmonsoon. Origin points are distributed over a larger area than in the pre-monsoon. Most post-monsoon storms originate in the northeast Arabian Sea in September and move gradually to the south-east and the south Arabian Sea in October–December. However, several storms formed in the Bay of Bengal and track west over India before arriving in the Arabian Sea. The tracks analysis shows that in the postmonsoon, cyclones usually track to the west of the Arabian Sea in September, then gradually to the south-west toward the Gulf of Aden and the Horn of Africa in October–December. However, some November cyclones recurved to India in the east of the Arabian Sea.

The origins and tracks for the sub-set of systems that made landfall in Oman also vary by season and month of formation. **Figure 1**(1) shows pre-monsoon origins and tracks, revealing a distinct difference in tracks within this period. In May, systems frequently travel to the coast between Masirah Island (central Oman) and Salalah in south-west Oman. In June the track direction moves to central to northeast Oman, such that tropical systems arrive at the coast between Masirah Island and Ras Al Had, the easterly most point of Oman at the mouth of the Oman Sea. Note, however, that the historical record indicates that storms occasionally deviate from this general pattern. A strong storm of May 1898 crossed Oman from Ras

#### **Figure 1.**

*Seasonal distribution of Arabian Sea cyclone tracks with landfall in Oman, 1881–2019: (1) pre-monsoon, (2) post-monsoon.*

Madrakah (south of Masirah Island) and moved to north Oman [25], whilst a storm of June 1885 moved to the south-east coast and entered the Gulf of Aden off Yemen [30].

density in mid-Oman, but additionally to the north of Oman, near Ras Al Had. In the post-monsoon, the highest density of the tracks is in mid-Oman near Ras Madrakah in September which moves to the south near Salalah in October.

*Kernel density estimation (KDE) of Arabian Sea tracks 1881–2019: (1) pre-monsoon, (2) post-monsoon.*

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard…*

*DOI: http://dx.doi.org/10.5772/intechopen.96961*

all storm tracks in the Arabian Sea (**Figure 4**(1)) the LDM is to the north-west toward Oman and Iran, with directional mean angle 127.8° (clockwise from due east; compass angle 322.25°), and an average track length of 1480 km. **Figure 4**(1) shows that the LDM in the pre-monsoon is to the north, toward Pakistan, with mean directional angle 120.4°, the post-monsoon LDM is to the north-west toward

Linear Direction Mean (LDM) results are presented in **Table 3** and **Figure 5**. For

**Figure 5**(1) shows the LDM of those tracks that made landfall in Oman. This track's mean directional angle is 157.8°, to the mid-East Oman coastline south of Masirah Island, and with a mean length of 2169 km. The LDM of the pre-monsoon tracks moved slightly to the north of the LDM of all tracks, to the middle of Masirah Island, with mean directional angle 146.0°, and mean length 1827 km. The postmonsoon landfall tracks have a mean directional angle of 157.9° to Ras Madrakah,

**Figure 5**(3) shows the LDM of tracks that made landfall in Oman, by seasonal distribution (pre-and post-monsoon), and **Figure 5**(4) shows the monthly LDM of tracks. May and June in the pre-monsoon are the highest frequency storm months.

*3.2.3 Linear direction mean of Arabian Sea storm tracks*

with a mean length of 2361 km.

**151**

**Figure 2.**

Oman's north-east coastline, with mean directional angle 136.4°.

**Figure 1**(2) shows that the origins and tracks in the post-monsoon similarly have a distinct tracks pattern within this period. In September cyclones track to Central Oman from Masirah Island to Ras Madrakah and then in October–December move progressively toward Salalah in south-East Oman.

#### *3.2.2 KDE of Arabian Sea tracks*

Kernel Density Estimation [KDE] is used to analyze cyclone tracks' distribution in the Arabian Sea and identify areas with a high-density of cyclones. **Figure 2** displays the KDE of all tracks in the Arabian Sea and shows a high density of tracks over a large area of the Arabian Sea, from 15–25° N and 60–73° E. In the pre-monsoon (**Figure 2**(1)) there is a high-density of track movement to the north-east, toward the north-west Indian coast (Gujarat), whilst in the post-monsoon (**Figure 2**(2)) the high-density area is more to the south and south-east of the Arabian Sea.

**Figure 3** shows the KDE of those tracks that made landfall in Oman. There is a high density of tracks in the mid-East coastline of Oman near Ras Madrakah, which is also evident in the pre-monsoon period (**Figure 3**(1)). In the post-monsoon period, the highest density of tracks is toward the south-east coastline of Oman, near Salalah (**Figure 3**(2)).

**Figure 4** shows the KDE of cyclone tracks making landfall in Oman, by pre-and post-monsoon month. In May, the tracks' highest density is in mid-Oman near Ras Madrakah, with some tracks in the south near Salalah. In June there remains a high

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard… DOI: http://dx.doi.org/10.5772/intechopen.96961*

**Figure 2.** *Kernel density estimation (KDE) of Arabian Sea tracks 1881–2019: (1) pre-monsoon, (2) post-monsoon.*

density in mid-Oman, but additionally to the north of Oman, near Ras Al Had. In the post-monsoon, the highest density of the tracks is in mid-Oman near Ras Madrakah in September which moves to the south near Salalah in October.

#### *3.2.3 Linear direction mean of Arabian Sea storm tracks*

Linear Direction Mean (LDM) results are presented in **Table 3** and **Figure 5**. For all storm tracks in the Arabian Sea (**Figure 4**(1)) the LDM is to the north-west toward Oman and Iran, with directional mean angle 127.8° (clockwise from due east; compass angle 322.25°), and an average track length of 1480 km. **Figure 4**(1) shows that the LDM in the pre-monsoon is to the north, toward Pakistan, with mean directional angle 120.4°, the post-monsoon LDM is to the north-west toward Oman's north-east coastline, with mean directional angle 136.4°.

**Figure 5**(1) shows the LDM of those tracks that made landfall in Oman. This track's mean directional angle is 157.8°, to the mid-East Oman coastline south of Masirah Island, and with a mean length of 2169 km. The LDM of the pre-monsoon tracks moved slightly to the north of the LDM of all tracks, to the middle of Masirah Island, with mean directional angle 146.0°, and mean length 1827 km. The postmonsoon landfall tracks have a mean directional angle of 157.9° to Ras Madrakah, with a mean length of 2361 km.

**Figure 5**(3) shows the LDM of tracks that made landfall in Oman, by seasonal distribution (pre-and post-monsoon), and **Figure 5**(4) shows the monthly LDM of tracks. May and June in the pre-monsoon are the highest frequency storm months.

Madrakah (south of Masirah Island) and moved to north Oman [25], whilst a storm of June 1885 moved to the south-east coast and entered the Gulf of Aden off

*Seasonal distribution of Arabian Sea cyclone tracks with landfall in Oman, 1881–2019: (1) pre-monsoon,*

progressively toward Salalah in south-East Oman.

*3.2.2 KDE of Arabian Sea tracks*

near Salalah (**Figure 3**(2)).

**150**

**Figure 1**(2) shows that the origins and tracks in the post-monsoon similarly have a distinct tracks pattern within this period. In September cyclones track to Central Oman from Masirah Island to Ras Madrakah and then in October–December move

Kernel Density Estimation [KDE] is used to analyze cyclone tracks' distribution in the Arabian Sea and identify areas with a high-density of cyclones. **Figure 2** displays the KDE of all tracks in the Arabian Sea and shows a high density of tracks over a large area of the Arabian Sea, from 15–25° N and 60–73° E. In the pre-monsoon (**Figure 2**(1)) there is a high-density of track movement to the north-east, toward the north-west Indian coast (Gujarat), whilst in the post-monsoon (**Figure 2**(2)) the

**Figure 3** shows the KDE of those tracks that made landfall in Oman. There is a high density of tracks in the mid-East coastline of Oman near Ras Madrakah, which is also evident in the pre-monsoon period (**Figure 3**(1)). In the post-monsoon period, the highest density of tracks is toward the south-east coastline of Oman,

**Figure 4** shows the KDE of cyclone tracks making landfall in Oman, by pre-and post-monsoon month. In May, the tracks' highest density is in mid-Oman near Ras Madrakah, with some tracks in the south near Salalah. In June there remains a high

high-density area is more to the south and south-east of the Arabian Sea.

Yemen [30].

*(2) post-monsoon.*

*Agrometeorology*

**Figure 1.**

**Tracks Compass**

*DOI: http://dx.doi.org/10.5772/intechopen.96961*

Tracks making Oman

landfall

**Table 3.**

**Figure 5.**

**153**

**Angle <sup>1</sup>**

*angle* ½ � *θ<sup>R</sup> of the resultant vector R*½ � *value up to 180° counter clockwise from due east.*

*Linear direction mean (LDM) parameters for Arabian Sea storm tracks, 1881–2019.*

**Direction Mean angle<sup>2</sup>** ½ � *<sup>θ</sup><sup>R</sup>*

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard…*

All tracks 322.25 127.75 0.28 67.72 16.82 1479 Pre-monsoon 332.59 120.41 0.24 67.04 18.96 1221 Post-monsoon 313.63 136.37 0.31 66.27 14.91 1569

Pre-Monsoon 302.25 147.74 0.08 63.47 17.11 1827 Post-Monsoon 287.39 162.60 0.03 64.71 16.25 2361 May 302.59 147.41 0.05 62.30 15.71 1748 June 306.30 143.70 0.12 65.35 18.91 1984 Sep 287.63 162.37 0.02 67.99 18.92 2424 Oct 289.22 160.78 0.03 65.87 15.45 2885 *Notes. 1. The resultant vector R*½ � *shifted into the correct quadrant clockwise from due north. 2. The directional mean*

*Linear direction mean (LDM) of Arabian Sea cyclone tracks: (1) all tracks, (2) tracks making landfall in*

*Oman, and (3) tracks making landfall in Oman by pre- and post-monsoon month.*

**Circular Variance** ½ �*s*

298.25 151.75 0.11 64.54 17.09 2169

**X Y** *OR*

**[km]**

#### **Figure 3.**

*Seasonal kernel density estimation [KDE] of Arabian Sea cyclone tracks 1881–2019, for cyclones making landfall in Oman. (1) pre-monsoon, (2) post-monsoon.*

**Figure 4.** *Monthly kernel density estimation [KDE] of Arabian Sea cyclone tracks 1881–2019 for cyclones making landfall in Oman.*

**Tracks Compass Angle <sup>1</sup> Direction Mean angle<sup>2</sup>** ½ � *<sup>θ</sup><sup>R</sup>* **Circular Variance** ½ �*s* **X Y** *OR* **[km]** All tracks 322.25 127.75 0.28 67.72 16.82 1479 Pre-monsoon 332.59 120.41 0.24 67.04 18.96 1221 Post-monsoon 313.63 136.37 0.31 66.27 14.91 1569 Tracks making Oman landfall 298.25 151.75 0.11 64.54 17.09 2169 Pre-Monsoon 302.25 147.74 0.08 63.47 17.11 1827 Post-Monsoon 287.39 162.60 0.03 64.71 16.25 2361 May 302.59 147.41 0.05 62.30 15.71 1748 June 306.30 143.70 0.12 65.35 18.91 1984 Sep 287.63 162.37 0.02 67.99 18.92 2424 Oct 289.22 160.78 0.03 65.87 15.45 2885

*Arabian Sea Tropical Cyclones: A Spatio-Temporal Analysis in Support of Natural Hazard… DOI: http://dx.doi.org/10.5772/intechopen.96961*

*Notes. 1. The resultant vector R*½ � *shifted into the correct quadrant clockwise from due north. 2. The directional mean angle* ½ � *θ<sup>R</sup> of the resultant vector R*½ � *value up to 180° counter clockwise from due east.*

#### **Table 3.**

**Figure 3.**

*Agrometeorology*

**Figure 4.**

**152**

*landfall in Oman.*

*landfall in Oman. (1) pre-monsoon, (2) post-monsoon.*

*Seasonal kernel density estimation [KDE] of Arabian Sea cyclone tracks 1881–2019, for cyclones making*

*Monthly kernel density estimation [KDE] of Arabian Sea cyclone tracks 1881–2019 for cyclones making*

*Linear direction mean (LDM) parameters for Arabian Sea storm tracks, 1881–2019.*

#### **Figure 5.**

*Linear direction mean (LDM) of Arabian Sea cyclone tracks: (1) all tracks, (2) tracks making landfall in Oman, and (3) tracks making landfall in Oman by pre- and post-monsoon month.*

Storms in May track toward the mid-east coast of Oman and Ras Madrakah. In June storms have a similar LDM but move to the north-west Arabian Sea and toward Oman's capital, Muscat, as they originate further north and east than May storms, whilst having a track length several hundred kilometers greater. A similar situation occurs in the post-monsoon. The LDM of tracks is similar for September and October storms, with October storms originating to the northeast of May storms. September storms thus tend to track toward Masirah Island on Oman's eastern coast, whilst in October, storms making landfall tend to head toward the south-east coast, and Salalah, Oman's main southern city.
