Terrain Analysis

By:
Dr. Archana Gupta

1
• Terrain is the lay of the land. This is usually
expressed in terms of the elevation, slope,
aspect and orientation of terrain features.
• Surface analysis is often referred to as terrain
(elevation) analysis when information related
to slope, aspect, viewshed, hydrology, volume,
and so forth are calculated on raster surfaces
such as DEMs (digital elevation models)

2
• The terrain plays an important role in many
spatial models. In a GIS, the information
about altitude is stored in digital terrain
models.

3
 Terrain Analysis
• Slope (Landslide susceptibility)
• Aspect (Solar insolation, vegetation)
• Catchment or dispersal area (Runoff volume,
soil drainage)
• Flow path (Distance of water flow to point)
• Profiles, fence diagrams
• Viewshed (visibility)

4
Definition
In scientific literature there is no
universal agreement about the usage of
the terms:
• digital elevation model (DEM)
• digital terrain model (DTM)
• digital surface model (DSM)
DEM is often used as synonimous of
DTM, but in many cases it used as a
generic term for both DTMs and DSMs.
The most common representation of a
DEM is a raster, where the DNs
correspond to the average elevation
value of the area framed by the cell.

5
 Uses
• Terrain analysis in geomorphology and
physical geography
• Modelling water flow for hydrology
• Modelling soil erosion or mass
movement
• Creation of relief maps
• Rendering of 3D visualizations
• Rectification of aerial photography or
satellite imagery
• Climatology
• Urban planning
• Logistics and communications
• Mining

• ...and many more other GIS

applications
 Sources of elevation data

• Stereo photogrammetry from aerial surveys

– traditional: manually by a trained photogrammetrist
• contours of topographic maps, to be interpolated
– modern: by automatic stereo-correlation
• Stereo-correlation from optical satellite imagery
– SPOT (off-track stereoscopy)
– Aster DEM (along-track stereoscopy)
• Interferometry from radar data
– European Remote Sensing Satellite ERS (multi-pass)
– Shuttle Radar Topography Mission SRTM (single-pass)
• Lidar
– interpolation of a matrix of xyz points obtained using a laser beam

7
 Grid vs. TIN
• In a digital terrain model, altitude values are usually organized in one
of these data structures:
• Grid
– The structures of grids are similar to raster data sets.
– Each rectangular grid cell has the same size.
– The altitude is stored in each of these grid cells.
– This data structure is the most widely used structure, due to its simplicity
and easily implemented algorithm.
– However, the disadvantage of a grid is that density cannot be adjusted to
the complexity of the terrain. For that reason there are often too many
data used for the representation of simple terrains.
• TIN
– TINs consist of irregularly distributed points.
– The triangles are built up on these points.
– This data structure allows for the efficient storage of terrain information;
the triangulation allows a variable density and distribution of points.
– It can be adapted to more complex terrain (more data points at complex
terrains and fewer points in flat areas of the terrain).
– It is more complicated to implement these algorithms than to implement
algorithms based on raster data.

8
GRID Vs TIN

9 9
Information which can be derived from
digital terrain models

10 10
 Spatial / Terrain Analysis
• The operations for spatial analysis of continuous fields that are
represented by the regular square grid, where each attribute is
represented by a separate overlay, and each grid cell is allowed
to take a different, scalar value.
• The z attribute can represent any continuously varying
attribute or regionalized variable, such as levels of pollutants
in soil, atmospheric pressure, annual precipitation, an index of
marketing potential, population density, or the costs of access
to a given location.

11
Basic operations for spatial analysis
with discretized continuous fields
• Map algebra and cartographic modelling

• Major advantage with raster representation in which each

attribute is recorded in a separate overlay, is that any
mathematical operation performed on one or more attributes for
the same cell can easily be applied to all cells in the overlay.
• This means that one can use exactly the same algebraic notation
to operate on gridded data as on single numbers. The method is
called Map Algebra and the procedure of using algebraic
techniques to build models for spatial analysis is called
Cartographic Modeling

12
 Spatial operations

• Using gridded data has advantages and disadvantages

compared with a topologically linked vector database of
defined entities.
• The disadvantages include the problems that the exact shapes
of entities are only approximated by the grid cells and that
directed operations over a network cannot be carried out
without first deriving the topology from the properties of the
surface.
• The advantages are that the continuous field model provides a
much richer suite of truly spatial analysis operations that have
many practical uses.

13
 These spatial operations include:
• Interpolation
• Spatial filtering
• First and higher-order derivatives
• The derivation of surface topology: drainage networks and
catchment delineation
• Contiguity assessment (clumping)
• Non linear dilation (spreading with friction)
• Viewsheds, Shaded relief, and irradiance.

14
 Interpolation
• Interpolation is the prediction of a value of an attribute z at an
unsampled site (x0) from measurements made at other sites xt
falling within a given neighbourhood. Interpolation is used to
create discretized continuous surfaces from observations at
sparsely located points or for resampling a grid to a different
density or orientation as in remote sensing images.
• Interpolation can be seen as a particular class of spatial
filtering where the input data are not necessarily already
located on a continuous grid.

15
Spatial analysis using square windows
• Spatial filtering
• The simplest and perhaps most widely used method of spatial
filtering a discretized, continuous surface involves passing a
square window (otherwise known as a kernel or filter) over the
surface and computing a new value of the central cell of the
window
• Ci,j as a function of the cell values covered by the window.
This kind of operation is also commonly known as
convolution.
• The window is frequently of size 3x3 cells, but any other kind
of square window (5x5, 7x7 cells, or distance measurements)
is possible.

16
• The most commonly used window operations are low- and
high-pass filters.

• Smoothing (low-pass) filter The value for the cell at the centre
of the window is computed as a simple arithmetic average of
the values of the other cells

17
18
• The low-pass filter has the effect of removing extremes from
the data, producing a smoother image. For nominal and ordinal
data the mean can be replaced by the mode, which is the most
common value.
• Using a modal filter on nominal data (e.g. soil units) can be a
useful way of simplifying a complex map but note that
smoothing a gridded image with a modal filter is a different
kind of operation than the procedure of generalizing a map by
reclassifying the attributes and merging the soil polygons.

19
 High-pass and edge filters
• The inverse of the low- pass filter is one that enhances the
short range spatial properties ofthe continuous surface,
enhancing areas of rapid change or complexity. The high-pass
filter is defined as:
• Original surface - Low-pass image = high-pass image

20
• A commonly used set of weights is called the Laplacian filter:

Edge filters are also used to enhance relatively uniform areas

in the continuum provided by a remotely sensed image. The
derivation of sharp edges and sets of boundary pixels is often
used for inferring the presence of discrete spatial entities in the
image which ultimately could be extracted and vectorized as
required.

21
 First and higher order derivatives
of a continuous surface
• Because the surface has been discretized, the derivatives are
approximated either by computing differences within a square
filter or by fitting a polynomial to the data within the filter.
• The two first order derivatives are the slope and the aspect of
the surface;
• the two second order derivatives are the profile convexity and
plan convexity
• Slope is defined by a plane tangent to the surface at any given
point and comprises two components namely, gradient, the
maximum rate of change of altitude, and aspect, the compass
direction of this maximum rate of change.

22
• Gradient and aspect are sufficient for many purposes, being
the first two derivatives of the altitude surface or hypsometric
curve,
• But for geomorphological analysis, the second differentials
• 1. convexity—the rate of change of slope expressed as plan
convexity and profile convexity—and
• 2. concavity (i.e. negative convexity) are also useful.

• Gradient is usually measured in per cent, degrees, or radians,

• Aspect in degrees (converted to a compass bearing),
• while convexity is measured in degrees per unit of distance
(e.g. degrees per 100 m).

23
 TERRAIN MAPPING
• Common terrain mapping techniques include
contouring, vertical profiling, hill shading,
hypsometric tinting, and perspective view.

24
 Contouring
• Contour lines connect points of the contour
interval represents the vertical distance
between contour lines, and the base contour
is the contour from which contouring starts.
• Suppose a DEM has elevation readings ranging
from 743 to 1986 meters. If the base contour
were set at 800 and the contour interval at
100, then contouring would create the
contour lines of 800, 900, 1000, and so on.

25
26
 Vertical Profiling
• A vertical profile shows changes in elevation along a
line, such as a hiking trail, a road, or a stream.
• The manual method usually involves the following
steps:
• 1. Draw a profile line on a contour map.
• 2. Mark each intersection between a contour and the
profile line and record its elevation.
• 3. Raise each intersection point to a height
proportional to its elevation.
• 4. Plot the vertical profile by connecting the elevated
points.

27
Automated profiling follows the same procedure but substitutes the
contour map with an elevation raster or a TIN.

28
 Hill Shading
• Also known as shaded relief
• hill shading simulates how the terrain
looks with the interaction between
sunlight and surface features.
• A mountain slope directly facing
incoming light will be very bright; a
slope opposite to the light will be
dark.
• Hill shading helps viewers recognize
the shape of landform features.

29
• Four factors control the visual effect of hill shading.

• The sun’s azimuth is the direction of the incoming light,

ranging from 0° (due north) to 360° in a clockwise
direction. Typically, the default for the sun’s azimuth is
315°. With the light source located above the upper-
left corner of the shaded-relief map, the shadows
appear to fall toward the viewer, thus avoiding the
pseudoscopic effect.

• The sun’s altitude is the angle of the incoming light

measured above the horizon between 0° and 90°.

• The other two factors are the surface’s slope and

aspect: slope ranges from 0° to 90° and aspect from 0°
to 360°
30
• Computer-generated hill shading uses the relative
radiance value computed for every cell in an
elevation raster or for every triangle in a TIN
(Eyton 1991).
• The relative radiance value ranges from 0 to 1;
when multiplied by the constant 255, it becomes
the illumination value for computer screen
display.
• An illumination value of 255 would be white and
a value of 0 would be black on a shaded-relief
map.

31
• The relative radiance value of a raster cell or a TIN
triangle can be computed by the following equation:
• Rf= cos (Af — As) sin (Hf) cos (Hs) + cos (Hf) sin (Hs)

• Where
• Rf is the relative radiance value of a facet (a raster cell
or a triangle),
• Af is the facet’s aspect,
• As is the sun’s azimuth,
• Hf is the facet’s slope, and
• Hs is the sun’s altitude.

• Suppose a cell in an elevation raster has a slope value

of 10° and an aspect value of 297° (W to NW), the sun’s
altitude is 65°, and the sun’s azimuth is 315° (NW). 32
• The relative radiance value of the cell can be computed by:

• R,= cos (297 — 315) sin (10) cos (65) + cos (10) sin (65) = 0.9623

• the depressions look like hills in an optical illusion called the

pseudoscopic effect

• The cell will appear bright with an R, value of 0.9623.

• If the sun’s altitude is lowered to 25° and the sun’s azimuth remains
at 315°, then the cell’s relative radiance value becomes:

• R,= cos (297 — 315) sin (10) cos (25) + cos (10) sin (25) = 0.5658

• The cell will appear in medium gray with an R; value of 0.5658

• The incidence value of a facet can be computed by:

• cos (Hf) + cos (Af — As) sin (Hf) cos (Hs)

33
 Hypsometric Tinting
• Hypsometry depicts the
distribution of the Earth’s
mass with elevation.
• Hypsometric tinting, also
known as layer tinting,
applies color symbols to
different elevation zones

34
 Perspective View
• Perspective views are 3-D views of the terrain
• The terrain has the same appearance as it would have if viewed at
an angle from an airplane.
• Google Earth has popularized perspective views of the terrain.
• Four parameters can control the appearance of a 3-D view
• Viewing azimuth is the direction from the observer to the surface,
ranging from 0° to 360° in a clockwise direction.
• Viewing angle is the angle measured from the horizon to the
altitude of the observer. A viewing angle is always between 0° and
90°. An angle of 90° provides a view from directly above the
surface. And an angle of 0° provides a view from directly ahead of
the surface. Therefore, the 3-D effect reaches its maximum as the
angle approaches 0° and its minimum as the angle approaches 90°.
• Viewing distance is the distance between the viewer and the
surface. Adjustment of the viewing distance allows the surface to
be viewed up close or from a distance.
• z-scale is the ratio between the vertical scale and the horizontal
scale. Also called the vertical exaggeration factor, z-scale is useful
for highlighting minor landform features.
35
36
• to show building features
in 3-D, we can extrude
them by using the
building height as the z
value

37
 Slope and Aspect
• Slope: It represents the rate of change in elevation at a specific
location. It can be expressed as a percentage (100 times the ratio of
vertical rise to horizontal run) or in degrees (arc tangent of the rise-
to-run ratio). Slope is a critical measure in terrain analysis.
• Aspect: This is the directional component of slope, indicating the
direction a slope faces. It starts at 0° (north), moves clockwise, and
ends at 360° (north again). Aspect is often categorized into principal
directions (e.g., north, east, south, west) or eight directions (e.g.,
northeast, southeast, etc.) for analysis.
• Both slope and aspect are essential in geographic information
systems (GIS) for terrain mapping and analysis. They influence
various applications, such as environmental planning, agriculture,
environmental studies, agriculture, urban planning, disaster
management and infrastructure development.

38
 Slope
• The slope at a point is given by the tangent
plane. In order to calculate the slope, the
derivatives of the surface in the direction of
the x- and y-coordinate have to be known.
• Therefore, special methods are needed to
estimate the derivatives from elevation
models. The most common method in grid
based terrain models is called "finite
difference"

39 39
 Slope From DEMs

If we move from left to right, the slope starts out close to zero
because the elevation is very flat, as shown in the DEM. Then,
the land drops in elevation and the slope goes negative. As the
land flattens out, the slope comes back to zero and then starts
to go negative again as the land slopes off

40 40
 Slope
• Definition: Slope refers to the rate of change
in elevation at a specific point on a terrain.
• Measurement:
• Percent Slope:

• Degree Slope: Calculated using the

arctangent of rise/run.
• Significance:
• Determines the steepness of a terrain.
• Essential for understanding land usability
(e.g., agriculture, construction).

41
 Calculating Slope
• Method 1: Percent Slope:
• Formula: Percent Slope=RiseRun×100.
• Example: For a rise of 10 meters and a run of 50
meters:
• Percent Slope=10/50×100=20%.
• Method 2: Degree Slope:
• Formula: Degree Slope=arctan(Rise/Run).
• Example:
• Using the same rise and run values:
Degree Slope=arctan(10/50)≈11.3∘.
42
 Aspect

• Definition: Aspect refers to the direction

a slope faces, measured clockwise in
degrees from 0° (north) to 360°.
• Significance:
• Determines sunlight exposure, which
affects vegetation growth, soil moisture,
and temperature.
• Crucial for agriculture, forestry, and
habitat studies.
• Categories of Aspect:
• North, South, East, West.
• Intermediate directions: Northeast,
Southeast, Southwest, Northwest.
43
Slope and Aspect

44
 Three Methods for Slope and Aspect
• All three methods use a 3-by-3 moving
window to estimate the slope and aspect of
the center cell, but they differ in the number
of neighboring cells used in the estimation
and the weight applying to each cell.

45
• The first method, which is attributed to Fleming and Hoffer
(1979) and Ritter (1987), uses the four immediate
neighbors of the center cell.
• The slope (S) at C0 can be computed by

• where ei are the neighboring cell values, and d is the cell

size.
• The nx component of the normal (e1 — e3), or the
elevation difference in the x dimension.
• The ny, component is (e2 — e4), or the elevation difference
in the y dimension.
• To compute the percent slope at Co, we can multiply S by
100
46
 Calculate Aspect
• S’s directional angle D can be computed by:
• D = arctan ((e4 — ey)/(e; — e3))
• D is measured in radians, which is with respect to the x-axis.

• If S <> 0 then
• T=D x 57.296
• If nx =0
• If ny < 0 then Aspect = 180
• Else
• Aspect = 360
• Elself nx > 0 then
• Aspect = 90—T
• Else 'nx < 0
• Aspect = 270—T
• Else 'S = 0
• Aspect = —1 ‘undefined aspect for flat surface
• End If
47
 Horn’s algorithm
• Horn’s algorithm uses eight neighboring cells
and applies a weight of 2 to the four
immediate neighbors and a weight of 1 to the
four corner cells.
• Horn’s algorithm computes slope at C0 as

48
49
 Sharpnack and Akin’s algorithm
• Uses eight neighboring cells but applies the
same weight to every cell. The formula for
computing S is:

50
 Computing Algorithms for Slope and
Aspect Using TIN
• Suppose a triangle is made of the following three
nodes:
• A (x1, y,, 21), B (X2, V2, 22), and C (x3, 3, Z3)
• The normal vector is a cross product of
• vector AB, [(x2 — x1), (y2 — y1), (z2 — Z1)], and
• vector AC [(x3 — x1), (y3 — y1), (z3 — z1)]. And
the three components of the normal vector are:

51
• The following diagram shows a 3-by-3 window of an
elevation raster. Elevation readings are measured in
meters, and the cell size is 30 meters.
• This example computes the slope and aspect of the
center cell using Horn’s algorithm:n, = (1006 + 2 x 1010
+ 1012) — (1017 + 2 x 1019 + 1020) = —37
• ny = (1012 + 2 x 1017 + 1020) — (1006 + 2 x 1012 +
1017) = 19

52
• For comparison, S, has a value of 17.16 using
the Fleming and Hoffer algorithm and a value
of 17.39 using the Sharpnack and Akin
algorithm.
• Aspect has a value of 299.06 using the
Fleming and Hoffer algorithm and a value of
296.56 using the Sharpnack and Akin
algorithm.

53
• Suppose a triangle in TIN is made up of following nodes with their x, y, and
z values measured in meters.
• Node 1: x1 = 532260, y1 = 5216909, z1 = 952 Node 2: x2 = 531754, y2 =
5216390, z2 = 869 Node 3: x3 = 532260, y3 = 5216309, z3 = 938
• The following shows the steps for computing the slope and aspect of the
triangle

• nx = (5216390 — 5216909)(938 — 952) — (5216309 — 5216909)(869 —

952) = —42534
• ny= (869 — 952)(532260 — 532260)— (938 — 952)(531754 — 532260) =
—7084
• nz = (531754 — 532260)(5216309 — 5216909)— (532260 —
532260)(5216390 — 5216909) = 303600

• Because S <> 0 and n, < 0, aspect = 270 - T= 260.546.

54
55
56
 Deriving surface topology and
Drainage Network
• Drainage networks and streams, catchments (or watersheds),
drainage divides or ridges are important properties of real
landscapes that contribute to the understanding of material
flows. They can be built into a TIN or an altitude matrix by
direct digitizing, but they can also be derived automatically
from the altitude matrix
• The following steps are required when automatically deriving
drainage networks from altitude matrices.
• 1. Determine routing
• 2. Removal of pits

57
 Determine Routing
• The flow of material over a gridded surface is determined by
considering the direction of steepest downhill descent. There are
several algorithms for calculating this, called the D8 or 8-point pour
algorithm, slope-weighted algorithms, and stream-tube algorithms
• The D8 (deterministic) algorithm approximates the flow direction by
the direction of steepest downhill slope within a 3 x 3 window of
cells. This leads automatically to a discretization of flow directions
to units of 45°
• The D8 algorithm computes a new attribute of flow direction which
can take eight different directional values which can be expressed as
degrees or as numeric codes. A useful implementation is to indicate
the directions by the numbers given on the numeric pad of a
computer keyboard, so that a sink or pit is indicated by a ‘5’.
58
The resulting new grid overlay is called
the set of local drain directions or Idd.
Each cell contains a directional type integer of value FD (flow direction) where:

59
• The Rho8 (random) algorithm is a statistical version of the D8
algorithm which was introduced better to represent the
stochastic aspects of terrain. It replaces the Wf of 1/V2 for
diagonals by 1/(2—r) where r is a uniformly distributed
random variable between 0 and 1.

60
 Removal of pits
When a smooth continuous surface is approximated by a square grid it is
inevitable that some cells will be surrounded by neighbours that all have
higher elevations.
The problem with artefact pits is that they disrupt the
drainage topology and need to be removed to obtain a
continuous ldd net. They can be removed by cutting
through or filling up

Cutting through one or more layers of boundary cells

to find the next downstream cell requires enlarging the
size of the search window to find a cell or series of
cells of the same elevation or lower as the core (pit)
cell.
Filling up involves increasing the elevation of the core
cell until it is equal to one or more of its neighbours,
and then examining if the neighbour drains downhill
to another destination. If this does not happen the
61
elevation is increased again until a linkage is found.
62
 Viewsheds, shaded relief, and
irradiance
• Concerning the computation of the paths of light between a
light source on or above the DEM and its effect at other
locations.
• Line of sight maps
• Shaded relief maps
• Irradiance mapping

63
 Line of sight maps
• This is the simplest operation; the aim is to determine those parts of
the landscape that can be seen from a given point. Intervisibility is
often coded as a binary variable—0 invisible, 1 visible. The
collective distribution of all the ‘true’ points is called the viewshed.

• Intervisibility maps can be prepared from altitude matrices and

TIN’s using tracking procedures

64
 Line of sight maps
• The site from which the viewshed needs to be calculated is
identified on the DEM and rays are sent out from this point to all
points in the model. Points (cells) that are found not to be hidden by
other cells are coded accordingly to give a simple map

• Because DEM’s are often encoded directly from aerial photographs,

the heights recorded may not take into account features such as
woods or buildings in the true landform and the results may need to
be interpreted with care. In some cases the heights of landscape
elements may be built into a DEM in order to model their effect on
intervisibilty in the landscape.

• Estimating the intervisibility of sites is an important GIS application

in simulators for pilot training, the location of microwave
transmission stations, the appreciation of scenery, or the location of
forest fire warning stations.
65
 Shaded relief maps
• Cartographers have developed many techniques for improving the visual
qualities of maps, particularly with respect to portraying relief differences in
hilly and mountainous areas.

• The shaded relief map computed from an altitude matrix differs from aerial
photographs in many ways.
• First, the shaded relief map does not display terrain cover, only the digitized
land surface.
• Second, the light source is usually chosen as being at an angle of 45° above the
horizon in the north-west, a position that has much more to do with human
faculties for perception than with astronomical reality.
• Third, the terrain model is usually smoothed and generalized because of the
data-gathering process and will not show the fine details present in the aerial
photograph.

66
 Shaded relief maps
• All that is required are the estimates of the orientation of a given surface
element (i.e. the components of slope) and a model of how the surface element
will reflect light when illuminated by a light source placed 45° high to the NE.
• The apparent brightness of a surface element depends largely on its orientation
with respect to the light source and also to the material. Glossy surfaces will
reflect more light than porous or matt surfaces

• Shaded relief maps can be extremely useful by themselves for presenting a

single image of terrain in which the three-dimensional aspects are accurately
portrayed.

67
 Irradiance mapping
• This is the extension of the shaded relief principle to compute the
amount of solar energy falling directly on a surface.
• The sun is now not fixed in any one position in sky, but is allowed to
take a position according to the latitude, the time of day, and the day
of the year.
• There is a need to incorporate the effect of atmospheric absorption
on the amount of energy actually received, and also to model the
shadowing effect of terrain.

• Integrate the daily or monthly estimates of irradiance for a whole

season or year, and thereby to create a map that distinguishes sites in
terms of the energy inputs for plant growth, home heating, or rock
weathering.

68
 Irradiance mapping

• soil information and vegetation indices from classified satellite

images can quickly provide a detailed reconnaissance model of the
landscape which can be used for hypothesis generation and
fieldwork planning.

69
 Visibility analysis
• Digital terrain models can be used for many
engineering and planning applications. In this
unit, visibility analysis is discussed in detail. A
visibility analysis detects all points that are
visible from a given point based on a digital
terrain model.

70 70
 Calculation of visibility
• In 1 Dimension, in order to determine if there is visibility
between point P and point V, a line of sight is calculated
between the two.
• If there is a higher point lying on the line of sight, the target
point is not visible. Otherwise, the target point is visible.
• The easiest way to perform such a test is to calculate the
slope of the line of sight (LOS) between point V and point P
and to compare it to the slope of the line of sight between
point V and point I hor , lying in between point V and point
P.
• The vertical angles of the two lines can be compared.
• In the case that the angle of the LOS between V and P is
bigger than the angle of the LOS between V and # hor , P is
visible, Otherwise, P is not visible.

71 71
72 72
In a terrain model based on TINs, the altitude of the intersection points between
the connection line V-P and the triangles have to be calculated first. The altitude at
the intersection points S1 to S n can be calculated from the two corresponding
vertices of the triangle by performing a linear interpolation. At this point, the
algorithm discussed above can be applied to calculate the LOS on the established
profile.

73 73
• Determination of the profile between the observation point
V and the target point P in a grid based digital terrain
model (raster):
• The profile is constructed on the intersection points
between the connecting line of point V and point P and the
edges of the grid.
• There are two possible approximation algorithms to
retrieve the altitude
• One is to retrieve the altitude in regular steps along the
LOS. The corresponding altitudes of the cells, in which the
points are lying, are assigned to the points.
• The second algorithm :The LOS is converted into a raster
and then compared to the digital terrain model.

74 74
75 75
• Visibility analyses are applied in different fields.
• For example, it can be used to determine the
location of an observation tower. By using a
digital terrain model, different possible locations
can be examined in a short period of time.
• The use of visibility analysis for the calculation of
the expansion of electromagnetic radiation (e.g.
for the planning of mobile phone antennas).
• A GIS canbe used to find the optimal area for the
relocation of animals. This reduces the possibility
of being discovered by a predator.

76 76
 Hillshade
• Sometimes, the relief in topographic maps is shaded to get
a better impression of the third dimension. For such kinds
of relief maps (hillshade), the illumination source is defined
generally at an angle of 45° from the north-west.
• Even though this position is very unrealistic for the
northern hemisphere, it is known that this sun position
gives the best impression of relief in the third dimension.
• The brightness of the shadow of a given surface element
(either a grid cell or a TIN section) depends on the following
properties:
• Aspect and slope of the surface element
• Reflecting properties of the surface element

77 77
78 78
 Shading methods
• Grid
• (Horn 1981) proposes a method for relief shading of grid
based maps (Burrough et al. 1998).
• In the first step the slope p, q has to be calculated for each
grid cell in direction x (east-west) and in direction y (south-
north):

• In the second step the reflection can be approximated as

follows:

79 79
Where p' is the slope in the opposite direction to the illumination source.
Let's set the illumination source on 45° north-west:

The parameter a gives the possibility to choose the gray level for
horizontal surfaces, b can be used to change the gray level with changing
slope

80 80
• TIN
• Relief shading of TIN based maps is performed
similarly to the grid based method except that
in this case, the reflection is not calculated for
grid cells but for triangles.

81 81
 Flow Direction
▪ Useful for finding drainage networks and
drainage divides
▪ Direction is determined by the elevation of
surrounding cells
▪ Water can flow only into one cell
▪ Water is assumed to flow into one other cell,
unless there is a sink
▪ GIS model assumes no sinks

82
 Flow direction in a DEM

340 335 330 340 345

337 332 325 335 340

330 328 320 330 335

328 326 310 320 328

320 318 305 312 315

Flow directions for individual cells
83
32 64 128

16 Source 1
Cell

8 4 2

84
 Finding watersheds …

▪ Begin at a source cell of a flow direction

database, derived from a DEM (not from the
DEM itself
▪ Find all cells that flow into the source cell
▪ Find all cells that flow into those cells.
▪ Repeat …
▪ All of these cells comprises the watershed
▪ The resulting watershed is generalized, based
on the cell size of the DEM
85
Watersheds …
Once done manually …

Contour lines (brown)

Drainage (blue)
Watershed boundary (red)
86
 Flow accumulation
▪ The number of cells, or area, which contribute
to runoff of a given cell
▪ The accumulation function determines the
area of a watershed that contributes runoff to
any given cell

87
 Flow accumulation in a DEM

0 0 0 0 0

0 1 3 1 0

0 1 8 1 0

0 1 13 1 0

0 2 24 2 0
Flow accumulation for individual cells
88
 Flow direction in a DEM

340 335 330 340 345

337 332 325 335 340

330 328 320 330 335

328 326 310 320 328

320 318 305 312 315

Flow directions for individual cells
89
Flow accumulation as drainage
network

Drainage network as
defined by cells above
threshold value for
region.

90
Watershed
•An area that contributes flow to a point on the
landscape
Water falling anywhere in the upstream area of a watershed will
pass through that point.

•Many be small or large

•Identified from a flow direction surface

Drainage network
•A set of cells through which surface water flows

•Based on the flow direction surface

91
92
 Routing and Allocation
•Routing
–Finding the shortest path between any nodes in a
network
–Optimal path
•Each link in the net can also be assigned an impedance value
•Using an accumulated distance and the impedance factor
–Most efficient route can be found, rather than just the shortest
–Nodes can also be coded with stops and barriers
•Preventing movement and forcing traffic along another path
–Although routing can be done in raster, it is much easier
if performed in a vector system

93
 Routing and Allocation
•Allocation
–Process used to define the areal extent of services
areas
–Service areas are defined around a site
•Region is formed that includes a defined area
–Location/allocation model (optimizes network
efficiency)
•Technique for the evaluation of multiple facility locations
–Determining the configuration of facilities (location)
–Assigning demand for the facilities (allocation)

94
 Errors and Quality Control
• Data accuracy is often grouped according to thematic
accuracy, positional accuracy, and temporal accuracy but
errors in spatial data can occur at various stages in the process
from observation to presentation.
• Errors in perception (improper identification) can occur at the
conceptual stage.
• Errors and approximations in determining the geographical
location depend on surveying skills, the provision of hardware
(GPS satellites, laser theodolites, etc.) and the choice of map
projections and spheroids. Errors in the measurement of
attributes, due to variation in the phenomenon in question, the
accuracy of the measurement device, or observer bias can
occur during the recording of the primary data.

95
• Errors can creep in when data are stored in the
computer because too little computer space is
allocated to store the high-precision numbers
needed to record data to a given level of accuracy.

• Finally, in the visual presentation of results, users

can obtain erroneous impressions if the semiotic
language is not clear, if colours and shading are
inappropriate or if displays are too crowded, or if
it is just too difficult to get a clear result.
96
 Data Quality
• Accuracy
• Spatial Accuracy
• Temporal Accuracy
• Attribute Accuracy
• Conceptual Accuracy
• Precision and Resolution
• Spatial resolution
• Temporal Resolution
• Thematic Resolution
• Consistency
• Completeness

97 97
 Data Quality and Error
• Spatial data and analysis standards are important
because of the range of organizations producing and
using spatial data, and the amount of data
transferred between these organizations.
• There are several types of standards:
– Data standards
– Interoperability standards
– Analysis standards
– Professional and certification standards

98
• National and international standards organizations
are important in defining and maintaining geospatial
standards:
– Federal Geographic Data Committee (FGDC) which
focuses on the national spatial data infrastructure
(www.fgdc.gov)
– International Spatial Data Standards Commission which is
a clearing house and gateway for international standards
– Open Geospatial Consortium (OGC) which is developing
interoperability standards. Web Mapping Service (WMS)
standards are an example.

99
 Spatial Data Standards
• Data – measurements and observations
• Data quality – a measure of the fitness for use
of data for a particular task (Chrisman, 1994).
• It is the responsibility of the user to insure
that the data is fit for the task.
• Metadata – data about the data

100
100
 Spatial Data Standards
• Spatial Data Standards – methods for structuring,
describing and delivering spatially-referenced data.
• Media Standards – the physical form of the data
(CD/download etc).
• Format Standards – specify data file components and
structures. These standards aid in data transfer.
• Spatial Data Accuracy Standards –document the quality
of the positional and attribute accuracy.
• Document Standards – define how we describe spatial
data.

101
101
 GIS Is Not Perfect
A GIS cannot perfectly represent the world for many
reasons, including:
• The world is too complex and detailed.
• The data structures or models (raster, vector, or TIN)
used by a GIS to represent the world are not
discriminating or flexible enough.
• We make decisions (how to categorize data, how to
define zones) that are not always fully informed or
justified.
• It is impossible to make a perfect representation of
the world, so uncertainty is inevitable
• Uncertainty degrades the quality of a spatial
representation
102
102
Concepts Related to Data Quality
• Related to individual data sets:
– Errors – flaws in data
– Accuracy – the extent to which an estimated value
approaches the true value.
– Precision – the recorded level of detail of your
data.
– Bias – the systematic variation of the data from
reality.

103
103
104
104
 Concepts Related to Data Quality
• Related to source data:
– Resolution – the smallest feature in the data set
that can be displayed.
– Generalization- simplification of objects in the real
world to produce scale models and maps.

105
105
Resolution and generalization of raster datasets
106
106
Scale-related generalization

107
107
 Data Sets Used for Analysis
Must be:
– Complete – spatially and temporally
– Compatible – same scale, units of measure,
measurement level
– Consistent – both within and between data sets.
– And Applicable for the analysis being performed.

108
108
A Conceptual View of Uncertainty

Real World

Conception

Source Data, Measurements &

Representation

Data conversion and Analysis

109
Result
109
Uncertainty in The Conception of Geographic
Phenomena
Many spatial objects are not well defined or their
definition is to some extent arbitrary, so that people can
reasonably disagree about whether a particular object is
x or not. There are at least four types of conceptual
uncertainty
– Spatial uncertainty
– Vagueness
– Ambiguity
– Regionalization problems

110
110
 Spatial uncertainty

Spatial uncertainty occurs when objects do not

have a discrete, well defined extent.
• They may have indistinct boundaries.
• They may have impacts that extend beyond
their boundaries.
• They may simply be statistical entities.
• The attributes ascribed to spatial objects may
also be subjective.

111
111
 Vagueness (obscureness)
• Vagueness occurs when the criteria that
define an object as x are not explicit or
rigorous.
• For example:
– In a land cover analysis, how many oaks (or what
proportion of oaks) must be found in a tract of
land to qualify it as oak woodland?
– What incidence of crime (or resident criminals)
defines a high crime neighborhood?

112
112
 Ambiguity
Ambiguity occurs when y is used as a substitute, or
indicator, for x because x is not available.
• The link between direct indicators and the
phenomena for which they substitute is
straightforward and fairly unambiguous (soil
nutrients for crop yield).
• Indirect indicators tend to be more ambiguous and
opaque (wetlands as an indicator of species
diversity).
• Of course, indicators are not simply direct or
indirect; they occupy a continuum. The more
indirect they are, the greater the ambiguity.
113
113
Regionalization problems
• Regional geography is largely founded on the
creation of a mosaic of zones that make it easy to
portray spatial data distributions.
• A uniform zone is defined by the extent of a
common characteristic, such as climate, landform,
or soil type.
• Functional zones are areas that delimit the extent
of influence of a facility or feature—for example,
how far people travel to a shopping center or the
geographic extent of support for a football team.
• Regionalization problems occur because zones are
artificial.

114
114
Uncertainty in the measurement of geographic
phenomena

Error occurs in physical measurement of

objects. This error creates further uncertainty
about the true nature of spatial objects.
– Physical measurement error
– Digitizing error
– Error caused by combining data sets with
different lineages

115
115
 Physical measurement error

• Instruments and procedures used to make

physical measurements are not perfectly
accurate.
• In addition, the earth is not a perfectly stable
platform from which to make measurements.
Seismic motion, continental drift, and the
wobbling of the earth's axis cause physical
measurements to be inexact. (GPSing error,
remote sensing error)
116
116
 Any digitized map requires:
Considerable post-processing
Check for missing features
Connect lines
Remove spurious polygons
Some of these steps can be automated
Digitizing Error
A great deal of spatial data has
been digitized from paper maps.

117
117
Error caused by combining data sets with
different lineages
• Data sets produced by different agencies or vendors
may not match because different processes were
used to capture or automate the data.
– For example, buildings in one data set may appear on the
opposite side of the street in another data set.
– Error may also be caused by combining sample and
population data or by using sample estimates that are not
robust at fine scales.
– "Lifestyle" data are derived from shopping surveys and
provide business and service planners with up-to-date
socioeconomic data not found in traditional data sources
like the census. Yet the methods by which lifestyle data
are gathered and aggregated to zones or are compared to
census data may not be scientifically rigorous
118
118
Uncertainty in the representation of geographic
phenomena
• Representation is closely related to measurement.
• Representation is not just an input to analysis, but
sometimes also the outcome of it. For this reason, we
consider representation separately from measurement.
– The world is infinitely complex, but computer system are finite.
– Representation is all about the choices that are made in capturing
knowledge about the world

– Uncertainty in earth model: ellipsoid models, datum, projection

types
– Uncertainty in the raster data model (structure)
– Uncertainty in the vector data model (structure)

119
119
 Sources of Error in Spatial Data
• Accuracy of Content
• Measurement Errors
• Field Data
• Laboratory Errors
• Locational Accuracy
• Natural Spatial Variation

120
Factors affecting the quality of Spatial
data

121
122
 Factors affecting the reliability of
Spatial Data
• Age of Data
• Areal Coverage
• Map scale and Resolution
• Density and Observation
• Relevance
• Data Format, Data Exchange and Interoperability
• Accessibility
• Cost and Copyrighting
• Numerical Errors in the computer
• Rounding Errors
• Geographical Coordinates and Precision

123
 Errors Resulting from Rasterizing a
Vector Data / Map
• 1. Grid Cells are only Approximation

124
2. Mixed Pixels

125
Raster – The Mixed Pixel Problem

Landcover map –
Two classes, land or
water

Cell A is
straightforward

What category to
assign
For B, C, or D?

126
 Error in raster
• raster
- because of the distortions due to flattening, cells in a raster can never be
perfectly equal in size on the Earth’s surface.
- when information is represented in raster form all detail about variation within
cells is lost, and instead the cell is given a single value. largest share, central point
(f.g. USGS DEM), and mean value (f.g. remote sensing imagery)

Largest share

8 6
Mean value
6 6
8 7.5 6.33
6

6.29 6
8 6
8x(1/6)+6x(5/6)=6.33
8 6 8x(3/4)+6x(1/4)=7.5
Central point 8x(1/7)+6x(6/7)=6.29
127
127
 3. Vector to fine Raster
• Converting polygons from vector to raster representation when
using grid cells smaller than the polygons brings with it the
problem of topological mismatch when the smooth polygon
boundaries are approximated by grid cells.

128
 Errors associated with the digital
representation (raster to vector)-
Vectorization
• The representation of curved shapes depends on the number of
vertices used. Consequently, the relative error of digitizing
straight lines is much less than that resulting from digitizing
complex curves.

129
 Error bands around a digitized line: problems
FOR POINT-IN-POLYGON SEARCHES AND WHEN
COMBINING RASTER AND VECTOR DATABASES

130
Uncertainty in the vector data
structure
• Socioeconomic data—facts about people, houses,
and households—are often best represented as
points.
• For various reasons (to protect privacy, to limit data
volume), data are usually aggregated and reported
at a zonal level, such as census tracts or ZIP Codes.
• This distorts the data in two ways:
– First, it gives them a spatially inappropriate
representation (polygons instead of points);
– Second, it forces the data into zones whose boundaries
may not respect natural distribution patterns.

131
131
Map Representation Error
Map scale Ground distance, accuracy, or resolution
(corresponding to 0.5 mm map distance)
1:1,250 0.625 m
1:2,500 1.25 m
1:5,000 2.5 m
1:10,000 5m
1:24,000 12 m
1:50,000 25 m
1:100,000 50 m
1:250,000 125 m
1:1,000,000 500 m
1:10,000,000 5 km

132
132