G 424/524 GIS for the Natural Sciences
D. Percy
e-mail: percyd@pdx.edu

Due: Week 4

Assignment 2

Part One: Data Discovery

You will continue to use the data from Assignment 1, trying to add useful layers to your project file. You might be interested in the relationship between landslides and some other spatial variable such as bedrock geology. Let's assume you are! First you need to find out where data that is at a high enough level of resolution (scale) to be useful might reside on the web.

Try these sites:

Metro

Oregon GIS Service

US Geological Survey

US Census

Geospatial Clearinghouse Network

Diva-GIS

Can you think of some other potential sources (try googling "gis data download <something you are interested in here>" or similar queries)? Characterize the above sites (plus some more that you find on your own!) by usefulness to the Natural Scientist (as opposed to social science). Note what file formats are available, and what coordinate systems are used. Spend at least 2 hours on this outside of lab time. A table is a good way to present the data you collected above!

Part Two:Analysis.

Characterize the geologic units by how many landslides occurred in them. You can also analyze the average volume this way. Do this graphically as graphs and with Choropleth maps. You will need to do summary operations on the tables after you have performed spatial joins. Here is the data to start with (geology_clipped_and zipped) (geologic_unit descriptions), next week you will go download and process these data, for this basic analysis just use what I've supplied.

Steps:

-1. Start with a new project file. Leave the project file from Lab 1 as is, in case you need to make any changes.

0. Unzip the geology data and add to a NEW MAP. Use new_geol, as it is 1:500K (higher resolution).

0.5 Define Projection (Toolbox->Data Management->Projections and Transformations) for new_geol_clip to be the same as Landslides. Your data need to be in the same projection for most ANALYSIS tools to work. So even though they may display correctly, the analysis tools will fail.

0.75 Click Add Data and add your landslides data to your map.

1. Spatial Join geology to landslides:

Right click landslides in the legend (TOC); choose Joins&Relates; from the first dropdown choose "Join data from another layer based on spatial location", choose new_geol_dd_clip to join to, relationship Falls inside, name the output file and note the location.

2. Summarize number of slides per Geology type (similar to SQL Select Distinct)

Open the attribute table of the shapefile generated by step 1, right click column to summarize (ptype or recno), choose summarize, click ok to get just the count without any other stats, note the location and name of output table, add result table to map.

3. Open the summarized attribute table (keep track of the Source and Display Tabs!!!) and make a graph of the data:

Right click and open the table output from step 2, click on Options and choose Create Graph and follow Wizard options, right click on blue top bar to get options like export or save on layout.

Alternatively, export the data to Excel or Matlab and graph it there. You have much more control in a real graphing program.

4. Summarize number of landslides per polygon.

  • Right click new_geol in the legend, choose Joins&Relates, Join
  • Choose "Join data from another layer based on spatial location"
  • Choose the first option for type of join under Points to Polygons, optionally choose some extra summary fields, give your output a name and location.
  • To get Avg Volume per Geologic polygon, when joining Landslides to Newgeol click on the Average and Std_dev checkboxes
  • Rt-Click the layer you just created, choose Properties, Symbology tab, Quantities, Count field, play around with different ways of showing these quantities
5. Join the geology descriptors to your summarized table from Step 2
  • Rt-Click the summarized table from Step 2 (ls_join_geol, in my case)
  • Choose Joins&Relates, Join
  • Choose "Join attributes from a table"
  • Choose Ptype for the "field in this layer that the join will be based on"
  • Find units.csv in the Folder browser (yellow folder)
  • Choose Unit as the "field in the table to base the join on"
  • Keep all records
  • Include this as a table in your report...

For the write-up:

Include a basic introduction, and the location map from the 1st lab. Also, remember that table of websites you spent a couple of hours outside of lab working on? Include that, too.

Try to do a SIMPLE analysis of the data. You have quantitative (tables) and qualitative (maps) data. Thus, you can visually and, at least at a basic level, Numerically Characterize (mean, st dev, variance, etc using the Summarize Field tool) the data. Limit the discussion to volume and frequency (landslides per Ptype). You really can't do much with only these two layers. Why? Remember, we don't just do this for fun, we are looking for answers, what else do you need? Think like a scientist.

IMPORTANT: Be sure to include some introductory text, and an overview of the processing steps. Be sure to include figure numbers and captions, be sure to reference your figures in the text!!! Embed your figures directly in your report, not buried in an appendix. Include your location map from assignment 1. Have fun!!! Discuss what layers of data you might need to really analyze this problem, and why you can't with just this one extra data set.

Upload your report to D2L Lab 2 folder.

revised Oct 11, 2016, 12:45pm.