ANALYSIS USING A GIS

Summary

Higher Technological Institute of Tlatluquitepec

Alejandra Ramírez Marcos IIAS Sabatino
Spatial data manipulation

The GPS (Global Positioning System) technology was originally designed for purposes
military but quickly its applicability in the civil world was seen, mainly in
navigation and topography.

The system consists of three components:

Spatial component, made up of 24 satellites located about 20,200 km from the Earth and
they pass through the same place every 12 hours in such a way that it can be counted at any time

place and at any time with at least 4 of these satellites at an elevation angle
of at least 150. Each satellite carries several highly precise atomic clocks on board.
and constantly emit a characteristic signal from each satellite that contains among other things
things the position of it.

Control component, made up of a series of observation stations close to the

Ecuador is responsible for controlling the orbital position of satellites and calibrating and synchronizing.

the watches.

Users with a GPS receiver among activities such as maritime or terrestrial navigation,
hiking, topography, machinery control, etc.
Depending on the needs, there are three modes of using a GPS device:
Autonomous navigation with a simple receiver, the precision is 20 meters for
military users and 100 meters for civilian users. Used in navigation
marina.
Corrected differential positioning (DGPS) with precision from 0.5 to 5 meters
used in GIS, coastal navigation, vehicle positioning, etc.
Differential phase positioning with precisions between 0.5 and 20 mm, used
in machinery control and topography.

The positioning calculation is based on measuring the distance from the position of
each satellite to Earth. Since the signal emitted by the satellite includes the time it was
issued and the receiver knows the arrival time, the distance can be calculated as:

d = cΔt
where c is the speed of light.
A pseudosphere is generated in this way with its center at the satellite and a radius equal to the

measured distance. If there are three satellites, there are 3 pseudo-spheres whose
intersection generates a unique point which is the position of the receiver. However for
to obtain a measurement, at least four satellites are needed due to the differences
temporals in the reception of satellite signals.

The distances measured by a GPS receiver are subject to the following sources of error:
Atmospheric delays on the signal;
Errors in the clocks;
Multipath effect, the signal can arrive at the receiver bounced from some
reflective surface (water sheets, buildings) obtaining a measure of
incorrect distance;
Loss of accuracy due to satellites being too close together. A GPS receiver
can provide a measure of this loss both horizontally (HDOP) and
in the vertical (VDOP);
Selective availability, distortions deliberately introduced by the military
American to reduce the accuracy for civilian users or foreign countries.
The satellites of the GPS system also have a signal that allows
high precision positioning, but is only accessible to applications
military.
The DGPS technique allows to solve many of these errors achieving precisions of
2 to 3 meters. It is based on the connection of the receiver to a fixed reference receiver. This
last, since it knows its position, it can calculate the GPS signal error and
transmit it to the mobile receiver so that it can correct it.

This correction can be done in real time or in post-processing using programs that
They download from the internet the files with the errors and the times that have been measured.
these errors to correct the positions measured by the mobile receiver and store them in
another file.

Change of projections

The process of transforming the geographic coordinates of the spheroid into coordinates
plans to represent a part of the surface of the ellipsoid in two dimensions are
is known as projection and is the traditional field of study of cartographic science. The
the emergence of GIS and the possibility of combining information from different maps with
different projections have increased the relevance of cartography beyond the mere
map making.

The fundamental problem when approaching a projection is that there is no way at all.
to represent the entire surface of the ellipsoid on a plane without deforming it, the objective will be
to minimize, as much as possible, these deformations. Since the effect of the
the sphericity of the Earth's surface is proportional to the size of the represented area (and in
consequence on the scale), these problems only arise when mapping large areas.

When it comes to mapping small areas, such as a city, distortion is

despicable because flat coordinates are usually used, relative to an origin of
arbitrary coordinates and measurements on the ground. These representations are called
plans instead of maps.

When the distortion due to the sphericity of the Earth's surface is considered relevant.
it becomes necessary to find an equation that corresponds to each pair of geographic coordinates

assign a pair of flat coordinates so that the different elements and objects of
the Earth's surface can be represented on a map. The units in which they
express the length in these new coordinates will generally be the meter,
allowing, in light of the incorporation of UTM mapping to a GIS, the simple calculation
of length, area or volume variables of the cartographed elements expressed in
units of the International System.
These equations are of the form:
= 1 ( , )
= 2 ( , )

To obtain these equations, the portion of the Earth's surface is projected (figure 1)
what will be mapped onto a geometric figure (a cylinder, a cone, or a plane) that if
can be transformed into a flat surface without distortions.

The focus of the projection can be located at different points, resulting in different types.
of projections. In this way, we can classify projections based on the object
geometric used for projection (figure 2), it is then referred to as projections
cylindrical, conical, and azimuthal or flat.

In the case of cylindrical or conic projections, the figure encompasses the ellipsoid and, after
to develop it, the result will be a plan in which a part of the Earth is represented.
through a Cartesian coordinate system. In the case of planar projections, the
The plane is tangent to the ellipsoid at a point and therefore does not need to be unfolded.

A projection always implies a distortion on the represented surface, the goal of

Cartography is about minimizing these distortions by using the most appropriate projection technique.

suitable for each case.

The properties of the ellipsoid that can be maintained are:

Conformity. If a map preserves the angles that two lines form in the
earth's surface, it is said that the projection is conformal. The requirement for
Conformity means that the meridians and parallels intersect on the map.
at a right angle and that the scale is the same in all directions around
a point, any point. A conformal projection also maintains the
small polygon shapes. This is a fundamental property in
navigation.
Equivalence is the condition by which a surface in the projection plane
it has the same surface area as in the sphere. The equivalence is not possible without
considerably deform the original angles, therefore, no projection
it can be equivalent and compliant at the same time. It is convenient, for example, in
cadastral plans.
Equidistance, when a projection maintains the real distances between two points.
points located on the surface of the Globe (represented by the arc of Circle
Maximum that unites them.
 Figure 1

Figure 2

As can be seen in figure 1, the distortions are null in the line where the figure
The geometric touches the ellipsoid and increases as the separation between both increases.
Therefore, to minimize the mean error, secant planes are usually used instead of planes.
tangents. Thus, instead of having a single line of the ellipsoid tangent to the figure
we have two secant lines and the distances to them, and therefore the errors, never
They will increase a lot. Thus, another criterion for classifying projection systems would be in

secant and tangent projections.

Integration and modeling of spatial data

Spatial data integration models using GIS are categorized into

different groups based on the methodology used or the objective pursued. In
In general terms, a GIS model can be considered as the process of
combination of a set of input maps or layers to produce an output map

MapExit=f(Map)1Map2, Map3Mapn)

The function f has different forms and can be categorized into three types depending on
of the nature of the expressed relationship: (a) based on theories and principles of physics and the
chemistry, (b) empirical, based on observations of data (statistical or heuristic), or
(c) some kind of mixture between theory and empiricism, which gives rise to another classification of
theoretical, empirical, and hybrid models, respectively.

Models based on weighted sum

This method is simple, it is based on multiplying each input layer by a

weight or weighting factor that indicates its importance in the model, and the subsequent sum
from the modified layers. The result is a new layer with numerical values, whose
magnitude represents on a relative scale the favorability of each area. This
the approach allows for a flexible selection of areas of interest, and also enables,
take into account the spatial distribution of them in the selection process.

The application of the weighted sum method to multi-class input maps involves
the assignment not only of a weight to each input layer of the model but also
the establishment of a different weight for each of the classes that make up each
map. The combination function used in this case is represented by the
simple expression

where F is the score indicating the favorability estimated by the model; Pi is the
weight for the input map i, and Pcl ij is the score for class j of map i. The value
it depends on the thematic class that appears in each pixel or polygon of map i.

The weighted sum of multiclass layers is a method frequently used in the

spatial modeling with GIS, primarily due to its conceptual simplicity, easy
application and notable flexibility when incorporating relevant information into the process
of combination.

Models based on multiple logistic regression

Regression analysis is a well-known statistical method widely used in numerous fields.

scientific disciplines, used to explore the relationships between different variables about the
basis of observations of those variables. The method involves deriving a relationship
mathematics between a set of independent predictive or explanatory variables and one
specific dependent condition.

Multiple logistic regression allows identifying the important variables in prediction.

of the probability of the presence of a deposit, in which the presence or absence of the
Deposit is defined as a dependent dichotomous variable. The coefficients of
regression for each variable derived from experimental observations in the area of
study, are used as weights in an algorithm applicable to the GIS database, by
example, through map algebra. In this way, a map is obtained that shows in
Probability terms (interval [0,1]) the presence of a mineral deposit, equivalent
the IFM. Quantitatively, the dependency relationship between the occurrence and the set of
predictive variables are expressed as:

where presence/absence of deposits, X1...Xpare the set of variables or layers

exploration information (for example, hydrothermal alteration, distance to fractures,
geochemical anomaly, etc.), and B0...Bpthey are the coefficients derived from the regression
logistics. That is, D is the dependent variable and X1pthey are the independent variables.

Recovery, reclassification and generalization

This is a crucial component of GIS, as it greatly influences the capacity of

user interaction with the data and their possibilities to restructure them to solve
specific problems. Very variable datasets can be retrieved, for example
by specified areas or regions, by topics or classes, etc., and two can be used
separate recovery systems - one for map data (lines, points, polygons)
and another for non-cartographic data (attributes, etc.).
The recovery for analytical purposes is normally carried out in interactive mode. In addition
of the possibilities of approach or selection of the windows mentioned above, the
Recovery functions generally allow for quick querying or searching.

The main classification for information retrieval models is as follows:

• Classic models: Among which are the probabilistic, Boolean, and

vectorial, which we describe in detail in this document.

• Structural models: Among which overlap-free lists and the method of the
proximal nodes.

Main models:

Boolean Model: It is a simple retrieval model based on set theory

and Boolean algebra. Given its inherent simplicity and its neat formalism, it has received
great attention and has been adopted by many of the early bibliographic systems
commercial. Their recovery strategy is based on a binary decision criterion
(pertinent or not pertinent) without any notion of a measurement scale, without the notion of a
partial matching under the conditions of the question.

Vector Model: The vector retrieval model or vector space model proposes a
framework in which partial matching is possible, assigning non-binary weights to the
terms index of the questions and the documents. These term weights are used
to compute the degree of similarity between each document stored in the system and the
user question.

Probabilistic Model: The probabilistic retrieval model is based on the equating

probabilistic, given a document and a question, it is possible to calculate the probability of
that this document is relevant to that question.

Complex analysis

GIS offer a wide range of analysis possibilities that can operate with
certain spatial aspects of the data, with their non-spatial attributes or with a
combination of both. The user performs the analyses interactively. In the Figure...
they illustrate some analytical techniques based on grids. Since there are so many types of analysis
different, we simply present some of the most common.
A. Spatial analysis. In addition to the integration and fusion of overlapping sheets for
to produce new mapped surfaces for analytical purposes, other analyses
spatial encompass the optimal allocation of routes (network analysis), the
calculations of intervisibility, slope, and aspect, in addition to the digital modeling of
land, the improvement of the allocation of places, the analysis of surfaces of
trend, etc. For the majority of these spatial analyses, the format is optimal
mesh, but it is likely that with the rapid advances in programming of
software the differences in the data structure will soon stop having
importance.
B. Statistical analysis. They encompass the summary and description of simple data.
numerical, for example the calculation of means, modes, medians, etc., up to
more complex operations such as frequency analyses and the measures of
dispersion, for example the automatic spatial correlations, the correlations
by ranges and the analysis of the closest term, as well as the variation analysis
multiple.
[Link] a series of operations that can be performed on one or more
data strata. They can be a simple enumeration of characteristics (of the total
or by polygons), the measurement of linear or curvilinear distances between objects or to
along them, the calculation of areas, perimeters or volumes, the calculation of angles
and the recording of direction measurements. Very simple measurements can be
 derive much more complex data, such as surfaces of
frequency/accessibility, various cost surfaces and any matrix analysis
user-defined.

Overlaps

Whatever the software, a GIS is built with the overlap of information; this
the analysis technique was developed and applied to planning in 1969 by Ian McHarg,
In his book 'Design with Nature,' he articulated several layers of information (slope,
runoff, soil types, rock composition, erosion, etc.) to generate maps
compounds of "physiographic obstacles" in the selection of road transport routes.
Superposition of entities

The key elements in the overlap of entities are the input layer, the layer of
superposition and the output layer. The superposition function divides the entities of the layer
at the entrance where they are overlapped by the entities of the overlay layer. They
create new areas where the polygons intersect. If the input layer contains lines,
the lines divide where the polygons intersect them. These new entities are stored.
in the output layer, the original input layer is not modified. The attributes of the
entities of the overlay layer are assigned to the corresponding new entities
in the output layer, along with the original attributes of the input layer.

Below is an example of line overlay on polygon. The line

divide along the edges of the polygon and each of the resulting line entities has the
original line attributes plus the attributes of the polygon in which it is located.
Raster superposition

In the overlay of rasters, each cell in each layer refers to the same one.
geographic location. This makes it suitable for combining the characteristics of several layers
in a single layer. Generally, numerical values are assigned to each characteristic, which
which allows you to mathematically combine the layers and assign a new value to each cell
in the output layer.

Below is an example of a raster overlay by sum. Two

Input rasters are summed to create an output raster with values for each cell.
added.

This approach is generally used to classify attribute values according to the

adaptation or risk, and then combine them to produce a general classification for
each cell. A relative importance can also be assigned to all layers to create
a weight classification (the classifications in each layer are multiplied by the value of
weight of that layer before adding them to the other layers).
Operations with neighbors; proximity analysis; relief study; analysis of
adjacency

One of the most basic questions asked about a GIS is 'what is near what?'

Proximity tools can be divided into two categories based on the type of
input that accepts the tool: entities or rasters. The tool-based on
entities vary in the types of output they produce. Distance tools
Euclidean distances based on raster measure the distances from the centers of the source cells.
to the center of the target cells. Cost-distance tools based on
Rasters accumulate the cost of each cell crossed between the origins and the destinations.

For entity data, the tools found in the set can be used.
of proximity tools to discover proximity relationships. These
tools generate output information with tables or entities of the area
influence. Influence zones are generally used to delineate the areas
protected areas surrounding the entities or to show the areas of influence. It can
use the multiple ring buffer tool to classify the areas of
around an entity in close distance classes, moderate distance, and distance
large for an analysis. Influence areas are often used to trim the data.
to a specific area of study or to exclude the entities that are within
from a critical distance of something for greater consideration in an analysis.

Lines of
influence, points and
rings.

The areas of influence can be used to select entities in another class of

entity, or they can be combined with other entities using a tool from
superposition, to look for the parts of the entities that fall within the areas of influence.

The distance from each point in an entity class to the entity can be calculated.
of point or line closest in another class of entity. It will also be added the
Entity identifier and, optionally, the coordinates of the nearest entity and
the angle towards her.

Connectivity analysis

GIS tools are very useful in the design of connectivity routes, as

allow to propose and evaluate various variables in the identification of the areas with the best
conservation opportunities.

Connectivity in terms of structure can be understood as the spatial configuration

of different types of habitats and is known as the degree of physical connection between the
patches that make up a landscape, also defined by the number of functional links
between patches of the same type, where each patch is connected or not based on a criterion
of distance.

The fragmentation and connectivity of the landscape can be assessed from the analysis of
spatial patterns at the structural and functional level, which are a process that shows
how the distribution of coverage types changes over time, and includes the
methods based on the evaluation of point patterns and evaluation methods
surface patterns.

Point patterns are those that describe a stochastic process that generates
an accounting set of events about space, where from one or several patterns
information is being sought about the process that has generated them. Pattern analysis
of surfaces, evaluates spatially continuous phenomena, both types of patterns
they can be evaluated using different spatial statistics.

Likewise, the aim is to compare that pattern with a theoretical one and classify it within patterns.

existing and interpreting that pattern. Usually, this classification consists of determining
if a) the pattern has an aggregated behavior (the events occur close to each other
events); b) regular (or inhibition where events appear scattered; c) or random
 (events occur with equal probability at any point in space, with
independence of where other events are located.

Network analysis

To perform a network analysis, you must work with a network analysis layer, which
it contains network analysis classes and objects.

Network analysis classes are the containers for analysis objects.

red. They can be tables or entity classes.

Network analysis classes can be input classes, output classes, or

input/output.

Input classes provide information to the solver during the

Input class resolution process. Create all objects in entry classes and set
its field values before conducting an analysis.

Exit Class Output classes only contain results.

Input/output classes are a combination of the other two classes. Create the
Class of
Objects for input/output classes and write some values if necessary.
input/output
of field before the resolution process.
The objects of network analysis are the entities or records in a class of analysis.
of network and serve as input and output for the network analysis layers. To perform
An effective network analysis requires knowing what the objects of analysis are.
of networks and how to work with them.

Some network analysis objects are logs that only provide data.
tables for analysis. They are only found in the problem analysis layer.
for generating routes for vehicles. Others are entities that contain a
geographical location, in addition to tabular data. However, the majority of the
network analysis objects are network locations that have tabular data, a
geographical location and a network reference position.

Network locations

A network location is a type of network analysis object tied to the network; it is

more, its position on the network is an input data for the analysis. Some
network locations, such as a stop on a route, can only be located in a
discrete point in the network; others, like a line barrier that represents a
construction, may cover a part of one or more edges.

Network Analysis Layer Network locations

Stops
Route
Point barriers

Facilities
Service area
Point barriers

Nearest installation Facilities

 Network Analysis Layer Network locations

Incidents

Point barriers

Origins

OD cost matrix Destinations

Point barriers

Orders

Vehicle routing problem Deposits

Point barriers

Facilities

Location and assignment Demand points

Point barriers

Example of network locations: bus stops

To better understand the fields of network location, consider the following

example. Two points that represent the start and end of a route (1 and 2 in the following)
(graph) are loaded as stops on a route analysis layer. Although the points
they do not fall exactly on the streets, the closest position in the network is determined
and is stored in the form of numerical values in the four network fields of the
stops. Next, the route analysis is conducted to find the best
route. The resulting route starts from the position of the nearest network to the
stop1(pointa) and ends at the nearest point of the network to the
stop2(pointb).

From the above example, it follows that stops 1 and 2 are network locations.
that are associated with the network at points A and B respectively. The following table of

attributes show the network fields of the points.