168 
The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B4. Beijing 2008 
3- MATERIALS AND METHODS 
Every point distribution is the result of a certain spatial process 
at a given time and a given space. The distribution of points 
(archaeological sites) on the landscape may have various 
patterns. They may take the form of clusters or they may be 
dispersed in a consistent distribution or their distribution may be 
entirely random without any specific pattern. With spatial 
archaeology there are various methods for understanding the 
distribution of archaeological sites in landscape and thus it is 
possible to detect spatial pattern from the point distributions and 
changes in point patterns at different times. Quadrat Analysis, 
which is one of the most common methods used in archeology, 
was employed in this study. Quadrat Analysis is used when sites 
are measured in terms of point rather than their weights. This 
method evaluates and measures the changes of distributions of 
points (the sites of the study) in terms of density and the number 
of points in each quadrat. The density, which is measured in the 
quadrat analysis, is compared to a hypothetical random pattern 
in order to find out from which pattern it is derived. This 
comparison is carried out within a framework of a spatial 
statistic system and its outcome is to arrive at a pattern that 
shows how the sites under investigation have formed. At the 
beginning of the analysis, it is crucial to determine the number 
and forms of the quadrat analysis. For this reason, in 2005 and 
2006 survey seasons, we first overlay the study area with a 
regular square grid (10m. x 10m.), and count the number of 
points falling in each square. Using precision military global 
positioning system (GPS) receivers with real time 5 m accuracy, 
aerial photography, a sighting compass and landmarks on the 
horizon, we were able to survey entire grids and mark the whole 
desired archaeological sites. Another important point about the 
approach to analysis of this research was determining the 
number and size of the quadrats. The studies of (Grifith and 
Amrheim 1991: 131) indicate that the required size of the 
quadrats can be obtained by the following equation: 
Size of the quadrat= 2A/r (1) 
Where A represents the size of the area under investigation and 
r represents the points in the distribution. 
By adopting the above equation, it became clear that given the 
right size, a quadrat has a width of 2A/ r when the quadrats in 
question are selected in the form of a square. Therefore, given 
the above correlation, it is possible to perform this calculation 
when the quadrats are selected with the required size. The 
number of quadrats can be obtained through the correlation n = 
r /2. When the area under investigation was located within 
coverage of quadrats, some of the quadrats were lacking in any 
kind of archaeological sites whereas some quadrats which had 
one, two, three, or more sites were distributed within. Then, the 
frequency of the points within each quadrat was counted and 
their density was measured (table 1). 
impact on identifying the properties of the site. For instance, if 
we envisage that the observed distribution pattern tends toward 
clusters, then we will need to look into the factors of this 
phenomenon. But if we encounter this phenomenon where the 
sites are distributed without any specific pattern, this might 
show that the usual factors such as environment which affects 
the sites do not have any role here. In fact, other factors more 
than the above come into play in site distribution and dispersion 
(see the rest of the article) 
Number of sites in 
each quadrat 
Observed frequency 
0 
38 
1 
8 
2 
4 
3 
8 
4 
1 
5 
2 
6 
2 
7 
3 
32 
1 
Total 
67 
Table 1. Frequency distribution of 118 sites observedfrom the 
eastern shores of of Urmia Lake 
In order to see the difference between an observed pattern and a 
pattern whose basis is a random process, we can use a common 
method, namely Poisson Process (equation 2) which is a 
suitable backdrop against which random point pattern can take 
place in the form of numerical data or frequency data. 
-A x 
e A 
(2) 
Where e is the natural logarithm and x t is the factorial of x . 
To illustrate the difference between the observed amounts and 
the amounts obtained from Poisson process, a statistical and 
analysis system and K-S (Kolmogrov Smirnov) were employed. 
K-S is a statistical method which measures the differences and 
similarities in statistics in frequency distribution. In running 
K-S measurement, our (H ) hypothesis was that there is no 
significant difference between the two distributions or if a very 
slight difference is observed, this difference is seen either as an 
error of sampling or a chance happening (Table 2). 
4- RESULTS 
In table 1 the distribution of sites within the quadrats can be 
seen in such a way that 38 quadrats don’t show to have any kind 
of archaeological sites and 8 quadrats exhibit only one site 
within it. On the other hand, one quadrat contains 32 sites. A 
glimpse at the frequency distribution of the sites within the 
quadrats may reinforce the idea that the sites within the quadrats 
tend to form in clusters. Even though this conclusion-up to a 
point- can be borne out by site distribution analysis, real 
corroboration occurs when the degree of difference and 
similarity of the observed frequencies is gauged in a 
measurement system in the form of statistics with a theoretical 
distribution basis. The type of site dispersion pattern has a huge 
The study of distribution pattern of 118 archaeological sites in 
the eastern shores of Urmia Lake and which was conducted by 
the use of archaeological ground survey in 2006 indicated a 
clustered pattern for archaeological sites. It isn’t the aim of the 
present article to identify the correlation of distribution pattern 
and existing factors in the area, because understanding the 
correlation of site dispersion pattern and environmental and 
cultural factors in the region plus the correlation of their 
interaction necessitates collecting and analyzing more pertinent 
data which they are at the preparatory stage. Nevertheless, the 
analysis which has been run so far reveals that up to a certain 
measure site distribution pattern follows a clustering pattern.