Geographic Information
Systems (GIS), a diverse and fascinating aspect of the study of geographies,
and a field in its own right. GIS is often employed to aid researches,
planners, and geographers better understand the relationships between various
factors within regional or the global environment.

A specific application
of GIS is Spatial Analysis. This is undertaken by geographic analysists who are
in need of specific and detailed information concerning the link between two or
more features within a given territory, and want to create conclusions or
identify patterns from the results which may hold current or future uses or
applications.

Literature
Review

A tool within the
arsenal of a Spatial Analyst is the Point Pattern Analysis. Daniel Erven (2018)
states that Point Pattern Analysis is an integral function of GIS and Spatial
Analysis, which evaluates the patterns and distributions of a set of points
within a predetermined space.

Daniel Erven (2018)
reveals some commonly used terminology that is often employed by Spatial
Analysts to discuss the phenomenon’s in the results of a Point Pattern
Analysis.

If it is understood,
in its more simplistic definition, that Point Pattern Analysis is the
determination of the number of points within a space, than it can be further
understood that the most basic Point Pattern Analysis terminology will include
the use of Descriptive Statistics.

Descriptive Statistics
are the common statistical tools often employed in any research containing
numerical data, and are used to count the number of points, identify the mean
and medians of the set of points, and to discover the standard deviation of the
set of points.

Considering how Point
Pattern Analysis uses points, the terms Frequency and Density are critical in
any spatial study. Frequency refers to the occurrence or occurrences of a point
within a time period, and Density shows the number of points within a certain
space.

Typically mentioned
when discussing Point Pattern Analysis are the terms Random, Uniform, Clustered,
and Dispersed.

Sets of points which
appear to be distributed across a surface according to a pattern or logic are
understood as being Uniform. Random sets of points are those which appear to be
distributed with a given area without pattern or logic.

Tony E. Smith (nd) illustrates
the contrast between a set of points which is Clustered, and another which is
Dispersed. Clustered points are those which are crowded around small areas
within a certain environment. Basically, if it can be observed that in a
surface area, the points are clumped together in concentrations with empty
space between them, then the points are Clustered.

The opposite is true
for Dispersed sets of points, implying that if the points are evenly
distributed across the surface area, with no apparent concentrations, then the
set of points can be understood as Dispersed.

Manuel Gimond (2017)
informs that density can be measured using Quadrants. This is when a surface
area is divided into smaller regions. Then the density of each Quadrant is
calculated to reveal the average density of each region. The rational of using
this technique is to simplify the information for the observer, and make it
more easily understood.

Phonetically and
practically similar to Descriptive Statistics, Descriptive Spatial Statistics
are often used in Point Pattern Analysis. Manuel Gimond (2017) explains that
Descriptive Spatial Statistics focus on the Mean Center, Median Center, and the
Standard Distance.

The Mean Center is the
average of all the X coordinates of a set of points, and the average of all the
Y coordinates of a set points, which indicates the location of the center of
the entire set. In similar fashion to the Mean Center, the Median Center will also
notify the Analyst of which point is the most in the center, but with the use
of Euclidian distances.

The Standard Distance
measures how far away all the points are from the Mean Center and informs the
researcher of the average difference of these distances. This allows for an
understanding of how clustered or dispersed the points are from the center.

Research
Objectives

The purpose of this
research is to ascertain the existence of any relationships between the
locations of hotels with the locations of attractions in the Muscat Governorate
by using Point Pattern Analysis.

Observing the data
without conducting any analysis, the author of this research hypothesizes that
within the entire research area, that there is no correlation between the
placement of hotels and the locations of attractions.

However, there is a
high likelihood that certain Willayats will demonstrate a definite tendency for
the placement of hotels near the locations of attractions, such as in Muscat
and Mutrah, which are highly desirable places for tourists.

It should be noted
that the author of this paper is working under the assumption that an
attraction is defined as “Something that attracts or is intended to attract
people by appealing to their desires and tastes” (Merriam-Webster, 2018).

 Study Area

The concerned area of
Study is the Muscat Governorate, which is one of eleven governorates in the
Sultanate of Oman, and is the Governorate in which the Capital City of the
Country, Muscat, is located. The Muscat Governorate is divided into six smaller
administrative regions known locally as Willayats. These Willayats are, Seeb,
Bowshar, Al Amerat, Quriyat, Mutrah, and Muscat.

Data Used

The data was provided
by the Ministry of Tourism, which is the government body responsible for the
administration, regulation, and growth of the tourism industry within the
Sultanate of Oman. The Data was acquired through Mrs. Ruqaiya Al Habsi who was
teaching the GIS and Spatial Analysis university module at the time. Please see
data below.

Methodology
Used in the Analysis

It will be critical
for the purpose of this paper to use a Point Pattern Analysis technique labeled
the Average Nearest Neighbor, as well as the Directional Distribution, and the
Mean Center methods of analysis. To do this, the author has selected to use the
ArcGIS software.

Applying these
techniques to the Hotel and Attraction sets of points, the observance of any
correlations will be facilitated. For instance, if the hotel set of points is
clustered, that may serve as evidence for the possibility of a relationship
between the hotels and attractions (as it would be logical that they would be
clustered around the attractions).

The possibility can
further be investigated by observing if the Mean Centers of the two data sets
are in close proximity of each other. Moreover, by using the Directional
Distribution on both data sets, it will be clear if the graphics are
overlapping in a way that could indicate a link between the two data sets.

For the purpose of
clarification:

The Average Nearest Neighbor as described by Ruqaiya Al Habsi (2017), calculates
the distances between each point (for one set of points), and the point which
is closest to it. All these distances are averaged to reveal the Average
Nearest Neighbor. This tool indicates whether the points are clustered,
dispersed, random, or uniform, which is useful in discovering the behavior of
one set points. ArcGIS (2002) explains that the Directional Distribution tool creates a
polygon which reveals the spatial characteristics of the set of points.

To be able to
determine the validity of the hypotheses, it is necessary to conduct the
research as such. Firstly, the author of this paper will apply these techniques
to the entire Muscat Governorate, in an attempt to discover if the first
hypothesis can be concluded as truthful or not.

Subsequently, it will
be necessary to apply the same principles to each individual Willayat to
discern the validity of the second hypothesis, which predicted that specific
Willayats display a pattern in the relationship between hotels and attractions.

 

Results

 

Muscat Governorate

For the entire
Governorate of Muscat, the Average Nearest Neighbor analysis reveals that the
Attractions are Clustered, with less than
1 percent chance that the result is due to random chance.

For the entire
Governorate of Muscat, the Average Nearest Neighbor analysis reveals that the
Hotels are also Clustered, with less than
1 percent chance that the result is due to random chance.

This provides a strong
possibility that both the Hotels and Attractions may be clustered in the same
locations. To determine the validity of this, it will be necessary to perform
the other spatial analysis functions.

 

Looking at the image
above, it can be seen that there is no apparent evidence to suggest that there
is a relationship between the Hotels and the Attractions in the Governorate of
Muscat. The Directional Distributions only overlap minimally, and the Mean
Centers are far from one another.

With this information,
it can be understood that Hotels and Attractions may share no apparent
relationships within the Governorate, and may both be clustered for other
reasons, such as being the result of transport networks, infrastructure,
natural phenomena, and so on.

There may however also
be some issue in the way the data has been analysed, for example, there are
many Attractions in the Wilayat of Quriyat but no Hotels, and these may serve
as outliers which strongly affect the analysis.

For this reason, it is
the authors firm belief that the individual Wilayats must also be analyzed to
reveal the possibility of any correlations between data sets.

Mutrah

From simple
observation, it appears that there may be some relationship between the Hotels
and Attractions in the Mutrah Wilayat.

The Average Nearest
Neighbor analysis reveals that the Attractions in the Mutrah Wilayat are Clustered, with a less than one percent chance
that this is the result of random chance.

The Average Nearest
Neighbor analysis reveals that the Hotels in the Mutrah Wilayat are also Clustered, with a less than one percent chance
that this is the result of random chance.

This serves as strong
evidence to support the hypothesis that certain Wilayats display strong
relationships between the Attractions and Hotels data sets. In the case of
Mutrah, with observation and the Average Nearest Neighbor analysis, it seems to
be clear that there is a pattern in the Wilayat.

Evaluating the
Directional Distribution and Mean Centers of the of the two data sets, it
becomes clearer that the relationship is not as strong as previously thought.
Although there is some overlap, and the Mean Centers are not so distant from
each other, the difference is considerable enough to come to the conclusion
that there is not a definite pattern between the Hotels and Attractions in the
WIlayat of Mutrah. This may be because hotels in the Ruwi area of the Wilayat,
which is a business and commercial hub, are established to cater to people who
want to do business in the area and not to visit attractions, as such there is
a lesser relationship from the two sets of points.

Bowshar

Upon first impression,
there appears to be a clear relationship between the Hotels and Attractions in
the Bowshar Wilayat, especially near the coastline.

The Attractions in Bowshar,
according to the Average Nearest Neighbor, are
Clustered. The analysis also reveals that the Hotels in the Wilayat
of Bowshar are also Clustered. This could
imply that there is a strong relationship between the data sets.

Looking at the Mean
Centers of the two data sets, it can be understood that there is a strong
correlation in the distribution of Hotels and Attractions in this Wilayat.
Perhaps the attraction far inland of Bowshar behaves as an outlier and
stretches the difference in distance between the Mean Centers, however it could
be concluded that the hotels are distributed in relationship the coastal
Attractions.

The Directional
Distributions of the two data sets show that the distribution of Hotels is very
similar to the distribution of the Attractions, only over a smaller surface
area. This could be, as before, due to the outlying Attraction in inland
Bowshar.

As such, the data infers a
positive response to the hypothesis, in which the conclusion that within the
WIlayat of Bowshar, there is a relationship between the Hotels and Attractions.

This heavily implies
that the Hotels in the coastal area of the Wialyat may be distributed around
the Attractions. This could further indicate that there is much visitor and
tourist interest in the area, and that any tourism development should
investigate the potential of this area.

Seeb

Looking at the distribution of
the two data sets, it appears that, as with the Bowshar Wilayat, there is a
relationship between the Hotels and the Attractions along the coastal areas of
the Seeb Wilayat.

 

The Average Nearest
Neighbor analysis reveals that both the Hotels and the Attractions within the
Wilayat are Clustered. This could be
indicative of a correlation between the two data sets, and serve as evidence
for a positive hypothesis.

Examining the
Directional Distribution and Mean Centers of both data sets, it is clear, that
there is without a doubt a relationship between the distribution of Hotels and
Attractions in the Wilayat of Seeb. The Directional Distributions overlap
almost exactly and are nearly the same shape and size. The Mean Centers are
within one kilometer of each other only.

With this evidence, it can be
confidently concluded that the majority of Hotels in the Wilayat of Seeb are
distributed in relation to the location of the Attractions.

 

 

Al Amerat

In the Wilayat of Al
Amerat there are two Hotels and no Attractions, as such it is impossible to
deduce any relationship between the Hotels and Attractions of this Wilayat, as
there can be no such relationship resulting from the fact that there are no
attractions.

Performing the Average
Nearest Neighbor analysis on the hotels of this Wilayat reveal that they are
spaced Randomly. As such, it can be
conclusively argued that the Hotels in Al Amerat are certainly not located
according to any pattern, and most definitely not spaced according to the
attractions of the Wilayat.

Muscat

Observing the
distribution of Hotels and Attractions in the Wilayat of Muscat, there seems to
be no clear correlation between the two data sets, with the exception of Sifa,
in the southern area of the Muscat Wilayat.

The Hotels and
Attractions in Muscat are both Randomly
dispersed, which reduces the likelihood that they are located in relation with
each other. This is not surprising for the Hotels, which can clearly be seen to
be located in a random pattern along the coast line of the Wilayat. However,
this is a strange revelation concerning the Attractions, which upon observation
would strongly suggest clustering, especially in the North of the Wilayat, in
the “City of Muscat”. This may be due to the outlying Attractions that have
distorted the analysis.

It appears that there is very
little similarity between the distribution of the two data sets. The only
common feature appears to be the direction of the Hotels and Attractions, which
are all in the coastal areas. There is a brief overlap in the Bandar A’Rowda
area, which may be due to the touristic or practical value of the area.

This lack of relation
between the location of Hotels and Attractions may also be due to the nature of
the Attractions. Most of the Attractions are located in the Old City of Muscat,
which houses many government buildings and the residence of His Majesty Sultan
Qaboos bin Said. As such, it may not be permitted to build or operate hotels in
that area, and as such, the hotels have been opened further south, or in the
adjacent Wilayat of Mutrah.

As far as the spatial
relation between the two data set, it can be said that there is very little or
no correlation between the Hotels and Attractions in Wilayat of Muscat.

Quriyat

 

As can be clearly
seen, there are Attractions in the Wilayat of Quriyat, but no Hotels. This
ensures that there is no relation between the Hotels and Attractions in this
Wilayat. In addition, the Average Nearest Neighbor analysis proves that the
Attractions are dispersed randomly, thus
reiterating the point that there can be no correlation between the two data
sets in this Wilayat.

 

Conclusion

This research provided
evidence that the first and second Hypotheses were positive. It could be seen,
in the Muscat Governorate, that although both data sets were clustered, there
was little justification to assume that they were clustered in the same locations.
Moreover, the observations supported the notion that certain Wilyats had a
stronger link between the Attractions and Hotels data sets.

The Wilayats where a
strong relationship between the Attractions and Hotels could be seen were Seeb
and Bowshar, and the Wilayats that displayed a lesser relationship between the
two data sets were Muscat and Mutrah. The Wilayats which displayed no
correlation were Quriyat and Al Amerat, as they only featured one of the data
sets.

The question may
arise, as a result of this research, as to why certain Wilayats display a
geographical relationship between Hotels and Attractions.

This may be the case
as a result of several reasons, such as the predisposition of certain Wilayats
towards the attraction of tourists or visitors, administrative, political, and
legal considerations, the availability of other nearby services, facilities, or
transportation networks (such as roads, ports, or airports), or simply the
difference in the level of development between areas.

To discover the truth
behind this question with absolute certainty, much more data collection would
be needed, and the relevant authorities would need to be interviewed to
determine what constitutes an attraction and how they are classified, what are
the limitations imposed by the Ministry of Housing to give away land, and what
are the restrictions imposed by the Ministry of Commerce and the Ministry of
Tourism for those interested in building hotels in certain geographic areas.

However, this paper
has been successful in presenting an interesting introduction on the
distribution of Hotels and Attractions in the Governorate of Muscat, and
gathering enough evidence to support the Hypotheses made at the formulation of
the research.