L12Interpolation With ArcGIS Pro
L12Interpolation With ArcGIS Pro
What You’ll Produce: maps, each with four panels. One map will have original
data and various sampled methods surfaces and the second map will show
interpolated surfaces.
We’ll apply both systematic and random sampling. We’ll also develop and apply
a stratification layer, because sometimes you want to stratify your sample, which
means you wish to increase sample density in some portion of your area, using a
map of zones, or strata.
1
GIS Fundamentals Lab 12
Add the DEM layer, chirdem, a region in southeast Arizona, to all four maps
and apply a color scheme to highlight the topography. (see examples below: Stretch,
Standard Deviation=1.5 Gamma =0.5; note I put for a 50% transparent Hillshade over Original just to make
more realistic; after you add chirdem & set the colors in the 1st map you can “copy” the layer & “paste” it into
the other 3; this approach saves steps)
Add a Layout (landscape) to your project, and place all four maps in the Layout, in
the pattern shown. Use the tools described in previous labs to assign a
1:300,000 scale and equal panel size to all four maps, at 3.5” wide by 3.1” high
(refer to previous lab, right click on a map square placed in the layout, then
properties).
Also specify a scale 1:300,000, you don’t need to type in the 1, :, or the commas.
Position and
label the Maps
in the layout so
that you have
something that
looks like the
figure
below/right:
Note that we
only added the
hillshade to the
Original Map,
and not to the
others.
We’ll now
create
sampling
points and
display them
with their
respective
maps.
Make sure the data frame windows DO NOT overlap, and the frames are equally-
sized and spaced, and the data in the frames are all at the same scale. Save
your Project now; you don’t want to lose all this set up work.
2
GIS Fundamentals Lab 12
Systematic Sampling
Activate your Systematic Map, switching to Map View
3
GIS Fundamentals Lab 12
Right click on the RASTERVALU column, and calculate the statistics as shown
before.
4
GIS Fundamentals Lab 12
Random Sampling
Activate the Random Map.
Generate a set of random sampling points via the geoprocessing tree, or the
geoprocessing tool search:
Tools ->Toolbox →Data Management Tools -> Sampling -> Create Random
Points
Calculate the mean value of RasterValu for your sampled points, and add it to
the title for the Random Map in your Layout.
5
GIS Fundamentals Lab 12
Sometimes we want to vary the sampling frequency across a map. Here, we’ll
place more samples in areas that are steeper. First we’ll create three zones, or
strata, and then we’ll assign samples based on these strata. Samples will be
assigned proportional to both the area of the strata, and the relative steepness,
with more samples in steeper strata.
Our strata boundaries will be based on slope, filtered to create larger, more
generalized areas.
Calculate the slope for chirdem (in the Toolbox, Spatial Analyst
Tools→Surface→Slope), specifying slope in degrees, and a z-factor set to 1,
(careful here, sometimes the Z factor is automatically set to a different value),
and saving to a permanent dataset named something like “slope_deg.” (this output
layer is not displayed in on this page)
Now to create strata. Usually you stratify for some threshold of an attribute, e.g.,
slopes above which you can’t build, or elevations where you’re unlikely to find a
resource of interest.
We’ll reclassify slope_deg into three classes; 1, 2 & 3 respectively using the
categories of 0-1.5, 1.5-18, 18 and larger. The Reclassify tool is in the
ArcToolbox, under Spatial Analyst Tools. Refer to the previous labs if you don’t
remember how to
permanently
reclassify, to a new
data set (Lab 11, page 13;
Lab 10, pages 15-16).
6
GIS Fundamentals Lab 12
Specify a FOUR as the Number of neighbors to use and the HALF replacement
threshold.
Notice how the small speckles are removed in the majority filter output, but the
long, thin reaches in valleys remain.
Again, this isn’t a great issue here, but in many cases, we want very general
strata, so we’ll apply another step.
7
GIS Fundamentals Lab 12
Notice how this substantially generalizes the data, removing the long, thin
features. (results of Focal Statistics are not shown on this page)
Convert the final FocalSt_Majo1 raster to a vector layer in the ArcToolbox via
Conversion Tools→From Raster→Raster to Polygon). You usually want to
simplify the polygons, and then here name it something like “Strata.”
Apply a symbology like that shown below. Inspect the layer’s data table. Note
there are Shape_Length and Shape_Area columns, these are in the projection
units, here meters and square meters. (Note: if you ever get “Table Failed to load data” when
you try an open an Attribute Table, remove the layer from the map and add it back to the map. The table will
now be relinked)
8
GIS Fundamentals Lab 12
Now we must create the stratified sampling points. (read to Page 12 BEFORE your try to
do start pushing buttons; you need to understand the process)
We would like to have a total of approximately 1000 sample points, with 10 times
as many sample points in the steep areas (red, at right) as in the flat (green
areas). We’d like three times as many samples in the intermediate areas
(yellow) as the flat areas.
The weightings are 10 for the steep, 3 for the intermediate, and 1 for the flat.
For example, in my layer above, the largest category, gridcode = 3 polygon has
an area of 255.5 square kilometers, and
the total area for category 3 is 272.3 square kilometers.
9
GIS Fundamentals Lab 12
So, this largest category 3 polygon should get 714 * 255.5/272.3, or 702 sample
points.
We multiply the number of points for the strata (category) by the polygon area,
and divide it by the total area of that strata.
or
Since we’ve already gone over the first two methods in previous labs, here we’ll
describe the Add Geometry Attributes geoprocessing tool.
10
GIS Fundamentals Lab 12
This case field means in all cases where the values for gridcode are the same,
sum them, e.g., sum all the areas for gridcode 1, do a separate sum for gridcode
2, and another for gridcode 3.
• 272.75 square kilometers for the flat (gridcode = 1) strata, and I want 72
sample points in this strata
• 182.35 square kilometers for the intermediate (gridcode = 2) strata, and I
want 214 sample points in this strata
• 217.93 square kilometers for the steep (gridcode = 3) strata, and I want 714
sample points in this strata
Your numbers may be slightly different, but should be with a few percent of these
areas if you used the methods we described above.
Write down your summary areas for all three of your classes, you will need
them later
1) Open the strata layer attribute table and add a new long integer field, call it
samples (see previous labs if rusty on how to do this: (note if Add if grayed out
check to see if you have “unsaved” edits, if so save them. If not save your Project and close ArcGIS
Pro and reopen it and reopen you Project.)
11
GIS Fundamentals Lab 12
4) Repeat this select calculate & process for strata 2 & 1, substituting the
strata area and total points appropriate for each strata.
To repeat, our general formula is:
POLY_AREA/(Total area of strata)
* (Total sample points for that
strata)
When you have finished with the last sample number calculation, be sure to
clear all selected features
12
GIS Fundamentals Lab 12
As with the other two Maps, Calculate the mean elevation with the Extract
Values to Points tool, and include the mean in the label for the stratified Map on
your Layout.
Add the stratified random sample layer polygon layer and the points to your
layout.
The map part of your layout should appear similar to the figure below:
Add a North arrow, legend, title, your name, and scale bar, create a PDF, and
turn it in on the course Canvas site.
13
GIS Fundamentals Lab 12
Interpolation
Add a new map, either in your existing project or in a new project, name the new
map Thiessen, and change the Map display coordinate system to UTM NAD83
(2011) zone 18N coordinates.
Display your data with graduated symbols, and with a topographic map
background, to appear something like the figure on the right. (Video:
InXY&Thiessen)
The spreadsheet data you loaded into OzGeog then to OzUTM, contains data on
ozone levels, a toxic element that may reach harmful levels, especially in the
summer, and in areas with large numbers of internal combustion engines.
The ozone column in the data set reports 24-hour average levels.
Readings higher than 54 are a moderate health risk, and readings higher than 70
are damaging to most populations.
14
GIS Fundamentals Lab 12
Readings are made at individual stations, but we’d like to create maps of the
concentrations across the northeastern U.S.
15
GIS Fundamentals Lab 12
16
GIS Fundamentals Lab 12
Geostatistical Wizard
We will now apply a series of interpolations using the interpolation techniques.
Create three new Maps, one labeled IDW, one labeled Kernel, and the last
labeled Kriging. (Video: GeoStatWiz)
Add the OzUTM data you used above to the IDW Map.
Many interpolation tools are provided through the Geostatistical Wizard, found
first by activating the Analysis tab on the main window (below left), then the
Geostatistical Wizard icon toward the right side of the tool ribbon (below right).
Double clicking on the Geostatistical Wizard (GeoWiz from here on) should
display a method selection window, with radio buttons next to various spatial
prediction methods:
17
GIS Fundamentals Lab 12
As you click each empty circle to choose a method, the parameters and
description of the method change. In the figure above, we’ve selected the
inverse distance method. Some of the names are different that those in the
textbook, e.g., the global polynomial method is the same as our trend-surface
method, and some of the methods are embedded within the choices, e.g., our
kernel interpolation is under their list of Interpolation with barriers, we would just
select this and apply without any barriers.
On clicking Next, you will likely get a warning saying multiple samples exist at the
same location. If so, select “Use Mean,” then Next
You should see a window similar to that above, with a preview interpolation
image on the left and a set of parameters on the left. Most should be familiar
from the textbook readings, except perhaps the Sector type and Angle options,
which allow you to apply the minimum/maximum neighbor thresholds to sectors,
rather than globally, e.g., with 4 sectors, you would allow 10 points each in each
of the northeast, northwest, southeast, and southwest directions, and constrain
the fits in each direction to match well at the edges of the quadrants. Sectors
18
GIS Fundamentals Lab 12
often provide better fits when there are underlying local trends, but details are left
for a more advanced class.
Vary the parameters, e.g., powers from 1 to 6, different neighborhood types, and
numbers of maximum and minimum neighbors, and observe how it changes the
preview surface.
The panel on the left side shows the predicted vs. observed values, a regression
line through the plot, and a 1:1 reference line of perfect prediction. You can also
click on the Error tab above the window that shows the error for each
measurement, the difference between the point on the plot and the 1:1 line.
The right side shows the cross-validation statistics (explained in the readings).
Note the RMSE of the cross-validation for this set of parameters.
Click on the back button, and only change the Power parameter, changing it from
1 to 2 on the preview interpolation screen. Then hit next again and view the
graph and cross-validation RMSE. How have they changed?
19
GIS Fundamentals Lab 12
20
GIS Fundamentals Lab 12
Activate the (empty) Map you created earlier and labeled Kernel, and add the
OzUTM data layer to it, symbolizing with proportional symbols as above.
21
GIS Fundamentals Lab 12
The next window shows the general options for the kernel estimator applied to
the ozone data.
Choose a set of
parameters and
estimate ozone
surface for the study
region.
22
GIS Fundamentals Lab 12
Conduct one final interpolation. Open/activate the empty Map you created and
labeled Kriging (or create it now).
Add the ozone data to your map and start the GeoWiz tool.
There are several variations on Kriging, and the options change substantially
depending on the Kriging methods you choose.
The covariance /
semivariogram
plot is
probably
the most
informative part
of this panel.
They show the
model fit in the
graph to the right,
and the
distribution of
binned samples
in the graph.
23
GIS Fundamentals Lab 12
Inspect the
distribution of
predicted and error
points, and note
the RMSE and
standardized RMS.
Click on finish to
estimate a surface.
Again, symbolize
the estimated
surface equal to
the previous
methods. (Manual, 3,
54;71;89, transparent,
60%, 60%)
24
GIS Fundamentals Lab 12
Add the vector shapefile STATES_GEOG and show only the state outlines (no fill
for polygons).
Set the Maps with the same scale and make each map box the same size (right-
click Map Properties, see previous labs for details).
Also include the usual elements: Title, name, legend, scale bar, north arrow:
25