Terrain analysis - Paul Bolstad at Work

By John Alexander,2014-03-28 02:43
8 views 0
Terrain analysis - Paul Bolstad at Workwork,Paul,paul

GIS Fundamentals Lesson 11: Terrain Analysis

    Lesson 11: Terrain Analyses

    What You’ll Learn: Basic terrain analysis functions, including watershed, viewshed, and profile processing.

There is a mix of old and new functions used in this lessons. We’ll explain the

    new, but you are expected to review old lessons if needed.

    You should read chapter 11 in the GIS Fundamentals textbook before starting.

Data are located in the \L11 subdirectory, all in NAD83 UTM zone 15 coordinates,

    meters, including driftless, a raster elevation grid, 3m cell size, Z units in meters, and driftless30, a raster elevation grid, 30m cell size, Z units in meters. A shape file called viewspot.shp is also used in this lab.

What You’ll Produce: Various hydrologic surfaces, a watershed map, a

    viewshed map, and profiles.


    Elevation data, also known as terrain data, are import for many kinds of analysis, and are available in many forms, from many sources and resolutions. In the U.S. there have long been available nearly nationwide data at 30 m resolution. Since the early 2000’s these have largely been replaced by 10 m resolution DEMs, and now many parts of the country are developing higher resolution DEMs, and 1 to 3 meters, based on LiDAR data collections.

    Although the most common use of DEMs is as shaded relief background for maps, we often are interested in working with terrain data calculating slopes,

    aspects, steepness or slope along profiles, viewsheds, as well as watershed and other hydrologic functions. The readings and lectures describe some of these applications, and this set of exercises introduces them.

    We will complete two projects here, using the same basic data. The first project has two parts, first comparing raster data sets of 2 vintages, a 30 meter data set from the USGS and produced in the late 1980s, and a 3 meter data set from LiDAR data collected about 2006. We then use the LiDAR DEM to explore profile and viewshed tools.

    Our second project focuses on watershed processing, using the ArcGIS Hydrology tools.


GIS Fundamentals Lesson 11: Terrain Analysis

    Project 1: Raster Surfaces, Profiles, and Viewsheds Start ArcGIS - ArcMAP

    Create a new map project, add the raster driftless to the view, and inspect it. Use the cursor and the layer Properties Source tab. What is the cell

    resolution? What are the highest and lowest elevation values?

    Now add the raster driftless30, and inspect it for resolution, cell size, and values.

Use the ArcToolbox;Spatial Analyst Tools;Map Algebra;Raster

    Calculator to subtract driftless from driftless30. Inspect the range of differences

    in elevation, and note where they largest differences tend to occur.

Unclutter the view by removing the driftless30 DEM and the difference layers

    from the map.

Derive the hillshade for driftless by

    ArcToolbox;Spatial Analyst

    Tools;Surface Analysis Hillshade,

    specify 25 for the Altitude and model

    shadows. Name the output file something

    like hs_drift.

Make the hillshade semi-transparent (Layer

    Properties; Display, then set the

    transparency to something like 50%, see

    Lab 6). You should get a display that looks

    approximately like the graphic to the right.

Turn on the 3-D Analyst extension,

    Customize;Extensions, then check 3D

    Analyst, then Customize;Toolbars; and

    also check 3D Analyst. (Video: Add profile)

    This displays a new set of tools, including the icons below.

    We’ll now explore the Line of Site and Profile View options. First, add the viewspot shapefile. This shapefile displays a single point, in the bottom left portion of the driftless DEM; it isn’t needed for a profile, but we will use it for our

    next operation.


GIS Fundamentals Lesson 11: Terrain Analysis

    Activate the Line of Site tool (). This will open a Line of site window (see right). Enter 2 for the observer offset (height, in meters here), and leave a value of 0 for the target offset (height).

    Now, left-click approximately on the point in the viewspot shapefile, found in the

     southwest quadrant of the driftless DEM.

    Move the cursor up to somewhere near the

    top of the DEM, and left-click again.

    If you make a mistake push Delete and

    start over.

This should display a profile line over the

    DEM, similar to the one shown on the


The profile line is a graphic element

    placed on the data and layout views.

    The default settings show areas not

    visible along the line colored in red, and

    areas visible colored in green.

Left-clicking on the Profile Graph tool displays the corresponding profile graph,

    as shown below.

    The profile graph shows the same profile

    line trajectory, but in a side view, and

    uses the same green/red coloring scheme.

    It shows the elevations on the vertical axis,

    and the horizontal distance on the

    horizontal axis.


GIS Fundamentals Lesson 11: Terrain Analysis

    Note that right-clicking on the profile graph displays a menu, and with the Properties option (see below)

    you may change the graph appearance, titles, and other properties.

    You may add a graphic of the plot to your layout by right clicking on the Profile Graph, and left-clicking Add to Layout (see below), or you may copy as a graphic or export to one of several image formats.

    After adding the graph to your layout, double

    check that it has appeared in the layout view.

    While the profile tool and graph show visibility

    along a specific path, we may wish to know all

    areas visible from a point, rather than just along a

    profile line. We could create a data layer that

    would show areas that are visible and hidden. We may do this with a viewshed function. To access this function via use: ArcToolbox ;Spatial Analyst Tools; Surface


    This should display the menu at right, in which you should specify driftless as the input raster,

    viewspot as the input point feature, an output raster name, and a Z factor of 1.

    This will then create a raster with all raster cells for the area covered by driftless categorized as either visible or not visible from the viewspot point.

    Make the non-visible cells transparent (assign no color to the symbols), and

    assign some appropriate color to

    the visible cells. Switch to the

    layout view, and add the usual title,

    name, north arrow, legend, and

    scalebar elements, similar to the

    graphic at right.

    Save your project, and export

    a .pdf of this layout, similar to that

    shown left.


GIS Fundamentals Lesson 11: Terrain Analysis

    Project 2: Watershed Functions

    Save and close your project, above, if you haven’t done so already. Then, open a new project, and add the driftless DEM.

We’ll be using ArcToolbox for this new project, but be forewarned these tools

    are buggy. You are best bet is to start early, follow the instructions carefully, save often, and if it doesn’t work, close ArcMap and restart from the beginning (meaning create a new project, and repeat the steps). If this doesn’t work, please contact one of the instructors for help no sense in banging your head

    against a wall. (see Video: Watershed).

Open ArcToolbox;Spatial Analyst Tools

    Hydrology. This should display the tools shown at the


We’ll be applying them in the following order:

    (The order is important.)


    Flow Direction

    Flow Accumulation

    Snap Pour Point (on a point feature we’ll create)


    We use the fill command to fill and pits in the DEM. As described in chapter 11 of GIS Fundamentals, pits are local depressions along which a stream would be expected to flow. These depressions may be real, or they may be due to errors in the data. However, a DEM is a representation of a waterless terrain surface, (the flow algorithms used here are programmed on this basis). So, local pits or flat areas may “trap” our routefinding when we are trying to identify downhill directions. The simplest of watershed processing routines begins by simply filling the pits. More sophisticated ones may fill the pits, and “burn in” a stream line, along which the DEM is lowered after filling to maintain a downstream flow.

Apply the Fill tool from the Hydrology toolbox. Specify the driftless DEM as

    input, and something like filled_dem for the output name. Don’t bother with the Z

    limit, here and in the subsequent tasks, and click on

    OK to run the tool.

After a minute, the filled data set should be

    automatically added to your view. Subtract (via the

    Raster Calculator) the driftless DEM from your

    filled_dem, and display the calculation. Change the

    color display and Display Background Value as

    dark blue. Notice the location and range of the fills;

     it should look something like shown on the right.


GIS Fundamentals Lesson 11: Terrain Analysis

    Now, apply the Flow Direction tool, using the filled DEM as your

    input, and specifying a flow direction output data layer; name the output f_direction. Check the box “flow all edge cells to flow


As noted in the textbook, this flow direction layer will contain a direction coding,

    with a set of numbers that define the cardinal and sub-cardinal direction, something like the figure to the right. The number 1 in a cell means water flows

    due east, 128 indicates northeast, 64 water flows north, and so on. This is a discreet categorization, all the water flows from a cell to an adjacent cell in only

    one of 8 directions. This method is known as the D8

    flow direction, to distinguish it from methods that can send a portion of the water from a cell to multiple

    neighboring cells.

Your output from the flow direction tool should look

    something like the figure to the right, and the

    symbols should show 8 values from 1 to 128,

    corresponding to flow direction.

Now apply the Flow Accumulation tool from

    ArcToolbox. This finds the highest points, and

    accumulates the area (or number of cells) downhill, according to the flow direction.

Specify the input as f_direction, name the flow accumulation raster as f_accum,

    ignore the weight raster, and specify an output data type of Float.

This should generate a display

    similar to the graphic at right. If you

    look closely, you will see some

    narrow, perhaps intermittent white

    lines in a dark background.

Notice the maximum and minimum

    value for the data layer, they will be

    something like 1.63 * 10^6 and 0.

These indicate the number of cells

    that drain to any given cell. For

    example, a value of 12,847 means a

    cell receives water from that many

    other cells. Since these cells are 3 m

    by 3 m, each cell counts for 9 square

    meters, and there are about 111,000

    cells per square kilometer.


GIS Fundamentals Lesson 11: Terrain Analysis

    We may use this flow accumulation grid to approximate where streams will be found on the surface, and to determine outlet points for watersheds.

    In any given small region there is usually a rough correspondence between drainage area, here measured with flow accumulation, and streams. For example, once a drainage area of 0.45 square kilometers is reached, a stream 222may form. This would be 0.45 km2 * 1,000,000 m/km / 9 m/cell, or about

    50,000 cells. So if we symbolize the flow accumulation layer so all cells above 50,000 are blue, and all equal or below are no color, we will get an approximate idea of where the streams will be found (this threshold is made for this exercise, but is probably not too far off here). Of course, stream initiation and location depends on many other factors, including steepness, soil types, vegetation, and precipitation amount and intensity, but this is a good first approximation.

     Reclassify the flow accumulation layer into two classes, and set the lower threshold to 50,000, the upper threshold to the maximum value, as described in previous labs. (Hint: Spatial Analyst Tools;Reclass;Reclassify, Classify use NoData for the

    under 50,000 values and 1 for above 50,000, name the output, d_streams)

    A few more steps will define our watersheds. First you must create an empty point shapefile (outlet.shp) using ArcCatalog, with the same coordinate system equal as the driftless DEM.

Then we must digitize a watershed outlet point. Add the new empty outlet

    shapefile, make it the active editing target, and display the DEM with the recolored flow accumulation

    grid over it.

Zoom into the southwest

    quadrant, so that your view

    looks similar to that at right.

    Digitize a point near the

    location shown below, along

    the stream near the outlet to

    the largest watershed in the


Remember, the figure at the

    right is zoomed to about ?

    of the DEM extent. You’ll

    want to zoom in much closer

    to digitize the point, so you

    can place it near the center

    of one of the stream cells.


    GIS Fundamentals Lesson 11: Terrain Analysis A larger scale zoom something like that shown at right is more appropriate. Start Editing, add your point, then save your edits and stop editing.

Now apply the Hydrology;Snap Pour

    Point tool.

; Specify Outlet as your input raster or

    feature pour point dataset,

; any pour point field, (Id is fine)

; the flow accumulation raster you

    created earlier, f_accum

    ; a new output raster to contain your raster pour point, name it r_pour_pt

; an appropriate snap distance, something like 3 to 9 (about 1 to 3 cells).

This should create a raster with a single cell for a value, near your digitized outlet.

Run the Hydrology;Watershed tool, specifying your flow direction raster

    (f_direction), your just created raster pour point layer (r_pour_pt), leaving the pour point field blank, and specifying an appropriately-named output watershed,

    like watershed.

    This should create a watershed layer something like that to the right.

    Changed the color, and make the

    raster 50% transparent.

Now convert the d_streams from a

    raster to a shapefile. This will display the streams more clearly.


    GIS Fundamentals Lesson 11: Terrain Analysis Use ArcToolbox;Conversions;From

    Raster;Raster to Polyline

    Name the output something like Derived_Streams.

    Select Background value as NODATA and Simplify polylines.

Create a layout displaying

    the driftless DEM,

    Watershed, Outlet and

    Derived_Streams with all

    the usual elements.


Report this document

For any questions or suggestions please email