Lab07_PatsCabin_DoDInputs.zip file. You will also want to download the DoD_Classed_2.5m.lyr file, because you will use the symbology from this layer to display your DEMs. Note: you don't need to add this layer file to your map.
You will also create a model that automates the process of filtering out the noise in the DEM of difference (DoD) that you create. More details below.
The reach has 13 beaver dams and/or support structures installed. After 2011 high flows, most of the top of reach ponds filled in completely with sediment. There was avulsion in the middle reach where significant downcut occurred as a result of the blowout of one of the major dams. The lower reach ponds aggraded substantially.
Model building is a way to automate your work. "When you create a model, you are
preserving a set of tasks, or a data processing workflow, that you can
execute multiple times." (ESRI)
You will build a model to automate the process of creating a DoD with the 20cm confidence threshold (steps 2-5 in the "A little more detail" section above). It might be helpful to run through the process so you know how the tools and inputs function. But you can go ahead and build the model first and use it to create a DEM of difference with the applied confidence threshold, if you feel comfortable with the process.
You can use the following video tutorials to acquaint yourself with model building in ArcGIS. The tutorial uses a different work flow (watershed delineation) but the process of model building is the same and should be pretty easy to follow. Again, you might want to run through the Task 3 steps of creating a DEM of difference so you are familiar with the inputs each tool requires. If you want to follow along with the example in the tutorial, you will need only 2 inputs: a
DEM and a pour point (provided in this
In this first video tutorial, we walk through creating a model that delineates a watershed:
In the second video tutorial, we go through how to make the model a more flexible tool that can be applied to any user input of a DEM and pourpoint:
On your Lab 07 website, display a captioned figure of your DEM of difference model (can be a screenshot), and provide a link to download the toolbox. (*.tbx) that you made.
Example screenshot of DoD Model
If you want a tutorial for creating a slightly more elaborate version of this tool, see here.
In Raster Calculator subtract DEM_PatsCabin_2010 from DEM_PatsCabin_2011
Evaluate your results. What are the range of values for this output raster? What are the units on these values? Does the range seem reasonable? Apply the color ramp (properties > symbology) "Cold to Hot, Diverging" to this initial DoD. Does the pattern make sense?
So if we change this threshold, how does it affect our results? Figure 2 shows a series of DoD outputs with varying levels of uncertainty removed. The image on the left is the raw output. The raster contains all values of elevation change. The image on the right has the largest threshold for uncertainty. (The grey values near 0 m have been removed and an underlying hillshade layer can be seen beneath the red and blue cells showing higher levels of erosion and deposition.)
The actual process of thresholding the DoD looks something like this:
The first step is to reclassify the cell values of the initial DoD to create a binary output raster where cell values of 1 equal areas of certain elevation change (greater than 20cm change), and 0 values being areas of uncertain elevation change (-20cm to 20cm).
In the Reclassify tool window, add your initial DoD and press the Classify... button to open the Classification window:
Change the number of classes to 3. If the window is greyed out, temporarily change the classification method to equal interval (ArcGIS glitch). In the Break Values section, change the breaks as shown above (you are setting the 2 breaks and leaving the max value alone). Click OK to return to the Reclassify window.
In the New values column, reassign the value "1" to all "true" cells (cells with certain elevation change) and assign NoData to the class of uncertain elevation change as shown below.
Set your output location and OK to run the reclassify. This output will essentially be used as a mask, as you will see in the next step.
The second step is to multiply your original (initial) DoD with the newly reclassified raster (values of 1 and NoData). This will create an output raster that preserves DoD elevation changes we are highly certain about and removes all elevation changes that are uncertain. If you multiply a cell value by NoData the output is NoData. Therefore the only elevation values that will be output will be in the areas we are certain about.
You are ready to symbolize your data in a meaningful way. In the final DoD raster's properties, go to the symbology tab and click to Show: "Classified" values on the left (see below). The symbology you want to import is classified, so you have to take this step for the import to work. Click the Import button and navigate to the Dod_classed_2.5.lyr you downloaded earlier.
Evaluate your results taking into consideration the following information about the site:
See What to Submit section on main Lab 7 page for specific submission details.