Title: Geography 360 Principles of Cartography
1Geography 360Principles of Cartography
2Outlines Isarithmic map
- Two kinds of isarithmic map
- Isometric map from the true point data
(continuous fields) - Isopleth map from the conceptual point data
(statistical surface) - Three interpolation methods
- Regular control points to gridded surface (e.g.
IDW, Kriging) - Irregular control points to triangulated surface
(e.g. triangulation) - Map display of interpolated data (2.5D map
display) - Vertical view contour, hypsometric tint
- Oblique view fishnet map, block diagram
- Physical model
Reading Slocum chapter 14 15
3What is isarithmic map?
- It depicts continuous smooth phenomenon
- Temperature, elevation, rainfall, average day of
sunshine, barometric pressure, depth to bedrock,
earths topography, and statistical surface
4- 1. Two kinds of isarithmic map
5Two kinds of data
- True point data
- Data is actually measured at the point location
- e.g. The location of weather station for
temperature map - This kind of map is called isometric map
- Conceptual point data
- Data is collected over areas, and the map is
constructed by interpolating given values at the
centroid of areas - e.g. The location of census tract for murder rate
map - This kind of map is called isopleth map
6Data types and isarithmic form
Dent 1999
7How isopleth map is created
Image source Electronic reading Nyerges
8Isometric or isopleth map?
- Think how data is collected
Current temperature in the US
Voting behavior in the US
9Isometric or isopleth map?
Toxic level
Demographic trends
Image source www.gis.com
10Depicting population distribution in different
map types
- Dot map total count in a large-scale mapping
- Proportional symbol map total count in a
small-scale mapping, where the point location is
conceptual (e.g. state centroid) - Choropleth map standardized data of population
(e.g. population density or cohort), where the
goal is to compare between enumeration units - Isopleth map standardized data of population,
where the goal is to reveal overall trends,
screening effects of arbitrary enumeration unit
boundary
11Phenomenon, data, mapWorld, data structure, map
displayElevation, DEM, shaded relief
- Toxic level map
- Phenomenon toxic level
- Data point data of toxic level at sample points
- Map isometric map showing continuous fields of
toxic level - Demographic trends map
- Phenomenon elderly persons ( population over
60) - Data point data of population density at the
centroids of enumeration units (the smaller the
better) - Map isopleth map showing statistical surface of
demographic trends
The process of transformation from point into
surface?
12- 2. Three spatial interpolation methods
13Spatial interpolation
- We will call the data points (either true or
conceptual) from which isarithmic maps are
constructed control points (not a common term,
but only for clarity purpose) - Spatial interpolation basically estimates unknown
values from known values at control points
guesswork generates a continuous surface from
sampled point values (which are discrete data)
because we know the phenomenon mapped is
continuous - Is it valid to apply spatial interpolation to
discrete phenomenon? - Figure 14.1 see how the manual spatial
interpolation works (it illustrates a linear
method) - Figure 14.4 see how different interpolation
methods yield different-appearing maps how can
we decide which method works the best given data?
14Location of weather stations
Surface map constructed from inverse distance
method
Surface map constructed from Kriging
15Spatial interpolation
- There are two ways to represent continuous
surface - one is a regular or gridded form, and
the other is an irregular form - Regular control points to gridded surface
- Inverse Distance Weighted (IDW) z f (h) where
h is distance to control points - Kriging z f (h, v) r where v is the
semivariogram model, and r is the residual (i.e.
difference between model and observed value) - Irregular control points to triangulated surface
- Triangulation z value is calculated from
Delaunay triangle
16Inverse Distance Weighted (IDW)
- As the distance increases, you will inversely
weight the values
Image source Bolstad 2005
See p. 274 for formula
17Kriging
- Similar to IDW in that
- A grid is overlaid on top of control points, and
the goal is to derive values at a grid point from
control points - Values at a grid are determined by values at
nearby control points weighted by inverse
distance - Different from IDW in that
- It builds the model of spatial autocorrelation
from known values (called semivariogram), and
the weights are determined such that observed
values are best fitted into the specified model - By model-fitting mechanism, the estimated values
are supposed to reflect the spatial structure of
given data it also provides the way to validate
the weights (e.g. standard error of the estimate)
18Triangulation
- Unlike IDW and Kriging, triangulation honors
control points - Triangulation helps determine the edge from which
values are interpolated - So how do we determine the best edges to work on?
- It works this way (see Figure 14.3B at p. 274)
- Draw Thiessen polygon from control points
- Thiessen polygon equally divides the area of
influence to each control point - Connect control points at neighboring Thiessen
polygons to result Delaunay triangle - Delaunay triangle minimizes the length of edges
formed by control points - Compare imaginary triangle IDE to ICE which is
smaller?
19Thiessen polygon
- Creates an area of equal influence given point
locations
20Discussion review questions
- Which spatial interpolation method do you think
can handle discontinuity? (e.g. lake as flat
plane instead of U-shaped gutter) - Which spatial interpolation method do you think
will produce inconsistent results depending on
parameters chosen (e.g. control points
considered) compared to others? - Which spatial interpolation method do you think
is considered a optimal method?
21Which method to choose?
- Triangulation honors the control point data, and
can handle discontinuity (e.g. ridge, lake) - Pros it works well when control points are
critical points - Cons angular contour
- Inverse distance fast, simplicity of method
- Pros easy to understand
- Cons deterministic method (no uncertainty
handling mechanism) - Kriging most rigorous method provided that the
model is properly specified - Pros stochastic method (uncertainty handling),
reflects overall spatial structure of data - Cons complexity of method, sensitive to model
specification
22Spatial interpolation in ArcGIS- Creating the
right surface map -
- Spatial Analyst
- Create contour from DEM
- Be aware of a wide array of parameters to choose
from - Geostatistical Analyst
- Provides exploratory tools for choosing the right
parameters or models, including cross validation
methods
23Image capture from Spatial Analyst
24- 3. Map display of interpolated data
252.5D vs. 3D phenomenon
- So far we have worked on 2D map display of
spatial entities (no height dimension) - Now we move on to 3D map display of spatial
entities - What we commonly refer to as 3D map display can
depict two categories as follows - 2.5D phenomenon (e.g. elevation) z value is
single-valued Color plate 14.1 depicts height
above a zero point - Z value is replaced by a single value of the
theme mapped - e.g. Prism map showing population density
- True 3D phenomenon (e.g. geological profile) z
value is multi-valued Color plate 4.1 depicts
geological materials underneath the earths
surface - Different values can be assigned to each (x,y,z)
- e.g. geological materials vary by (x,y,z)
Read Slocum p. 57
26Displaying the interpolated data
- Vertical view
- From Gods eye view 90
- Contour lines (Figure 14.16A)
- Hypsometric tints (Figure 14.16B)
- Hill shading (shaded relief) (Color plate 15.3,
Color plate 15.2C) - Oblique view
- From Birds eye view 090
- Fishnet map (Figure 14.16C)
- Block diagrams (Figure 15.17)
- Physical model
27Contour lines
- Each contour line depicts the same elevation
28Hypsometric tint
- Space between contour lines is color-coded
- Can be either classed or unclassed
29Hill shading (shaded relief)
- Illuminate earths topography with imaginary
light source
Cardinal direction of light source? What do you
think determines the reflectance values of pixel
in digital image?
30Which map display is this?
31Oblique view
- Fishnet map Block diagram
32Oblique view
- Fishnet map Fishnet map draped image
33Physical model
- 3D map representation of 3D