Thursday working session. Today has two halves: this deck, where we walk the site-selection workflow one more time and hand out the final project, and the concepts review deck for Exam 2. Have the final project instructions open on Learning Suite before class so students can follow along.
Working session, not a lecture. The goal is that everyone leaves with a chosen project, a chosen county, and a first draft of their processing steps written down.
Lab 11 (Walmart Site Selection) is the closest model for what the final project asks for. Walk it once more, out loud, before handing out the project.
Start with the plain-English conditions. Do not let anyone name a tool yet: first the conditions, then the data, then the tools. This is the same order the final project report has to follow.
The jump from "near a highway" to "within 2 miles of I-15 or a highway" is the whole trick. Every criterion needs a number and a layer before you can run a single tool. Their final project criteria have to be written this way too.
Point out that three of the four are downloads and the fourth they make themselves. Digitizing your own layer is a legitimate data source as long as you say so in the report. Note the agency renamed from AGRC to UGRC; older lab handouts and slides still say AGRC.
These all live in the QGIS Processing Toolbox under Vector overlay and Vector geometry. Erase is the one name that changes: in QGIS it is Difference. One more tool that is not an overlay: the Field Calculator, in the attribute table window, computes a new attribute column, e.g. density = population divided by area.
This is the figure from chapter 9 of the textbook and it is the single most useful picture in the course. The order matters: a different order gives a different final layer. Their report has to contain a diagram like this one, drawn for their own project.
A and C both work and give the same result; D is the trap. Order matters whenever an operation changes the extent that a later operation sees, and clipping early is usually faster because every later step runs on less data. Ask them to defend an answer before you say which. This is the same idea as the Figure 9-1 true/false question that shows up on the concepts exam.
The next seven slides are one long analogy. Run them quickly, one click each. Students remember this.
Rolled dough is the whole world. Before any criterion is applied, every location is still a candidate.
Line up the cutter: choose the layer that defines your study area.
The cut shape is a new layer. The dough that is left over is discarded. That is a Clip.
Now a second, smaller cutter: intersect the state with the county so only the county remains. Each cut narrows the candidate area.
Difference (Erase in ArcGIS): remove the water. You cannot build a store in the lake, so the lake polygon is cut out of the candidate area.
Buffer the existing stores by 2 miles, then Difference those buffers out. What is left is every location that satisfies all the criteria at once. Each hole is one store's exclusion zone.
The output is a brand new layer that did not exist before. That layer, symbolised and put on a map with the required elements, is the deliverable.
Emphasise "keep them". Students routinely overwrite their intermediates and then have nothing to show for the processing steps, which is a graded part of the report. Tell them to name layers as they go: step1_clip, step2_buffer, step3_difference, and so on.
An example final map made in QGIS 3.44 Print Layout: a small site-selection result (cell towers within 1 km of a major road in Utah County) with title, subtitle, legend, scale bar, north arrow, data sources, projection, a neat line, and the result stated in one sentence. This is the standard. Past student maps often add inset maps zoomed to the best candidate sites, which is a good addition when the candidates are small.
Read the full instructions document on Learning Suite with them; this slide is the summary, not the specification. The "write the steps down first" line is the one that saves them the most time.
Give them two minutes to pick before you walk the five option slides. Any of the five works with the same tool set they used in Lab 11.
Solar farms want flat ground, high solar radiation, a nearby transmission line and a road, and land that is not already built on or protected. Slope from a DEM, land cover, and transmission lines are the obvious layers. The linked video is a short piece on utility-scale solar siting; click the image to open it in a new tab.
Landfills are the classic exclusion problem: away from housing, away from streams and wells, away from floodplains, off steep slopes, on the right soils, but still within a reasonable haul distance of the population it serves. Almost all Difference and Buffer.
Airports need a long flat run of land, clear approach paths, distance from dense housing because of noise, and road access. Slope and land cover from raster data, plus population from Census blocks.
Reservoirs want a narrow valley with a large upstream drainage area, close to the farmland being served, and not on top of existing development. This one leans hardest on terrain data.
Temple siting is a membership-density and drive-time problem: far enough from existing temples, close to a large membership, visible, and on a parcel large enough to hold the building and its grounds.
The report is a short technical document, not an essay. Two to three pages of text is enough; the maps carry the rest. Anything they downloaded needs a source named.
The intermediate maps are where most of the grade separation happens. A single beautiful final map with no evidence of the processing is not a complete submission. Screenshots from the QGIS map canvas are fine for the intermediates; the final map goes through Print Layout.
One PDF. Not a folder of images, not a QGIS project file. Point them at the instructions document on Learning Suite for the grading breakdown.
TODO: the sign-up sheet link lives on Learning Suite; the 2018 deck linked a Google Sheet that is no longer current. Paste this term's sheet link here or drop it in the Learning Suite announcement before class.
Spend the last fifteen minutes of class on this. Walk the room; the students who leave without a county chosen are the ones who fall behind.
Set the expectation that work sessions are for getting unstuck, not for starting from zero.
Remind them the Testing Center has its own closing hours and that the exam is not open book.
Conversion notes (2026-09-02): source "Final Project Discussion.pptx" (2026, 25 slides). Nothing dropped outright; source slides 14 and 15 both used the same cookie photo and were merged into one slide, and the "Get Started" Google Sheet link was replaced with a pointer to Learning Suite because the 2018 sheet URL is stale (marked TODO). Software wording updated to QGIS throughout: ArcMap digitizing to QGIS digitizing, Erase noted as QGIS "Difference", ArcToolbox implied tools mapped to the Processing Toolbox, AGRC updated to UGRC. ArcGIS-era artwork kept and flagged: images/fp-example-map.jpg is a student final map produced in ArcMap (map output rather than a software screenshot, so it is still a fair example, but a QGIS Print Layout example would be better). The tool-order quiz slide and the QGIS name table are new, written for this session.
Update 2026-09-02: ArcGIS-era screenshots replaced with QGIS 3.44 captures made by tools/qgis_reshoot_screens.py: fp-example-map-qgis.png (QGIS Print Layout example) replaces the ArcMap student map.