Project 2 - Assignment & Deliverables

ENV 761 - Landscape GIS   |   Spring 2024   |   Instructors: Peter Cada & John Fay  |   
Due 16-Feb
After several conversations, WWF feels the products listed below will be hugely beneficial for their workshop. Please submit the following no later than 16-Feb. Submit them as hardcopy and also digitally on Canvas. ALSO, submit your ArcGIS Pro toolbox to Canvas. (We’ll use this to give you partial credit…)

  1. First, present the stream dataset you created from the 90m DEM along with a brief memo explaining any relevant decisions you made in creating the dataset.

    • Include a map set to the extent of the 5 watersheds showing the watershed boundaries and your stream layer. Show larger order streams with thicker lines.
    • Also include a small inset map zoomed in to an area showing your generated stream along with the INEGI stream layer. (And yes, I realize this isn’t the perfect layer to show as a background “truth” layer, but it’ll due here…)
    • Discuss in your memo how your DEM derived stream layer compares to INEGI’s in terms of accuracy and level of detail. Include how you might be able to increase or decrease the number of streams appearing in your derived dataset.

  2. Next, WWF-Mexico wants to know the tradeoffs between using the 90m DEM vs. the 15m DEM. Specifically, they want you to calculate summary statistics (mean and standard deviation) of slope (degrees) for each of the 5 watersheds. Submit the following for the slope data derived from both the 90m and 15m DEM.

    :point_right: NOTE:

    • Convert the watersheds created in Part II - Task 4.2 to polygons (simplifying the shape & creating multipart features) to use as zones for computing summary statistics.
    • Ensure that you are using the geodesic method in computing slope and you are computing values in degrees.
    • Be sure to round values appropriately; the nearest 10th of a degree is fine.

    • Enter the mean and standard deviation of slope for each watershed in the first table in the Project2_Results.xlsx Excel file (found here).

    • Discuss the results generated by the different resolution DEMs. Are slopes more or less the same, or does one appear to generate higher slope values than the other? Briefly explain why they are similar or different. (One paragraph will do here.)

      TIP: When migrating values from ArcGIS tables to Excel, if you save the table in .dbf format (by adding the .dbf extension to your file name and NOT storing it in a geodatabase), you can open it from within Excel. Or you can also export ArcGIS tables to CSV or to XLSX formats.

  3. Some researchers will be interested in your Slope Position and Landform datasets. Provide the following maps for discussion.

    • Provide two panel map of showing each of the two slope position datasets – fine and coarse. Mention the neighborhood values used to generate each and briefly explain how the two maps differ in terms of what’s being shown. Set the map’s extent to the 5 watersheds (same ones used above) and display the watershed boundaries. (Tip: Convert to polygons…)
    • Provide a map of your landforms, again zoomed to the 5 watersheds and showing their boundaries. Include a table listing how much area of each landform is included in the full dataset (i.e. not clipped to watershed boundaries, just the entire raster). Report areas in $km^2$, and round appropriately.

  1. As a very rough approximation of biodiversity within each watershed, WWF wants you to compute the relative amount of warm and moist area within the 5 catchments. They’ve agreed that “warm” areas are those with an insolation > 135 and moist areas are those with a TCI > 11.

    NOTE: Use the TCI dataset your derived from the D-Inf Flow algorithm as your proxy for wetness.

    • Given this criteria, calculate the total “warm and moist” area within each catchment. Insert these values in the second table in the Project2Results.xslx table.
    • Include a map showing the pixels classified as “warm” and “moist” zoomed to the extent of the 5 watershed boundaries. Include the watershed boundaries in this map.

  2. WWF wants to map sections of the landscape that may be more sensitive to deforestation, grazing, and human impacts. They suggest using the technique Weiss mentions in the last frame on his poster (link) which examines the level at which streams occur in deep incised valleys across a landscape. To do this, they’ve asked you for the following:

    • Create a TPI raster using an annulus with an inner radius of 5 cells and an outer radius of 10 cells. Use the 90m DEM.

    • Using the planning units dataset created in this lab (Part 3 - Task 6) and the TPI dataset just created, compute the mean TPI value for each planning unit.

    • Show the result as a map of the 364 planning units classified into 5 quantiles (i.e. quintiles) using the mean TPI attribute.

    • Now, extract just the pixels in this TPI dataset that fall within 500m from your 90m stream raster, using flow length not Euclidean distance. Then, create a second map showing the mean values of these near-stream TPI cells within each of the planning units. Again, display these values in quintiles and don’t forget to include legends!

      In the end you should have two maps that allow us to compare mean TPI values per planning unit - one for the entire unit and the other for just the areas near streams. Be sure that your map’s legend includes the quintile divisions. See the last panel in the Weiss poster for an approximation how this should look.

  3. Lastly, WWF is interested in the relative amount of floodplain occurring in each of the 5 catchments. They consider a floodplain to be areas that are up to 3 meters above the stream.

    • Create a map displaying the floodplain areas across the 5 catchments, using the floodplain derived from the 15m DEM as instructed in the lab PDF. (Again, use the catchments derived from the 90m DEM, but convert them to polygons so that cell size does not become an issue.)

    • Enter the total floodplain area (in $km^2$) occurring in each in the Project2Results.xlsx table.