What I'm researching about Clemson's water and trees before it goes on the main site. Six rooms, one question each. Click a room to walk in.
These rooms hold the research I'm working on before any of it goes on the main site. I'm trying to figure out what happened to the water around Clemson, and where we could start putting it back in the ground.
Each room asks one question. Walk them in order, or go wherever you want. Every claim on a board links to where it comes from.
Only sending someone one link? Send the watershed comparison. People get it fastest. The Clemson Water page ties the water work together.
When rain lands on a forest, most of it soaks into the soil and reaches the creek slowly. When it lands on a roof or a parking lot, gutters and pipes carry it straight to the creek.
So the first thing I wanted to know was how much ground around Clemson switched from the first kind to the second.
Sources: Walsh et al. 2005
I pulled the USGS land-cover maps for every year back to 1985 and counted.
Sources: USGS Annual NLCD
That shows up in the creeks. USGS researchers in Atlanta compared an urban creek with less-developed neighbors: in the biggest storms its peak flows ran 30 to 100% higher, and in dry spells its low flows ran 25 to 35% lower.
The water that used to soak in and feed the creek between storms leaves all at once instead.
Sources: Rose & Peters 2001; Peters & Rose 2001
The full page has the land-cover breakdown, a heat map, and a cross-section of a constructed wetland, one way to catch that water and let it soak back in.
Open A Water Story →One watershed doesn't tell you much. So I ran six of them through the same steps with the same public data, from Golden Creek at 1.5 square miles up to the Reedy River at 62.7.
Each one is a room off this hallway. On the blueprint, how full a room is tells you how much of that watershed is paved.
Sources: USGS Annual NLCD
The Center for Watershed Protection's impervious cover model draws the lines: streams start showing stress at about 10% paved, and past 25% most stream life can't hold on.
Paved means roof, road and parking, from 2023 land cover.
Sources: Schueler 1994; Schueler et al. 2009
Ecologists call the whole package the urban stream syndrome. It works the other way too: anything that slows water down and lets it soak in cools it, cleans it, and feeds the creek between storms.
Sources: Walsh et al. 2005
Drains to Lake Conestee / Saluda River. Clean Water Act: 11 impaired segments: impaired biota (cause unknown), metals, pathogens (bacteria).
Open Reedy River →Drains to Lake Hartwell (Seneca arm). Clean Water Act: Not individually listed.
Open Hunnicutt →Drains to Lower Eighteenmile Creek. Clean Water Act: Not individually listed.
Open Berkeley →Drains to Twelvemile Creek / Lake Hartwell. Clean Water Act: Not individually listed.
Open Golden →Drains to Coastal Plain (Moncks Corner). Clean Water Act: Not individually listed.
Open California Br. →Drains to Lake Hartwell (Seneca arm). Clean Water Act: 3 impaired segments: nutrients, pathogens (bacteria), turbidity.
Open Eighteenmile →Sources: Schueler et al. 2009; USGS Annual NLCD; NOAA Atlas 14; EPA ATTAINS
Across all six, an ordinary two-year storm now sends 482 acre-feet more water downstream than the same storm did on 1985 ground. That's about 157 million gallons.
Every bit of it can be slowed down, spread out and soaked in close to where it falls. The comparison page lets you try that with rain gardens.
Compare the six →Sources: NOAA Atlas 14; USGS Annual NLCD
A creek's channel is shaped by the floods it carries. Send it bigger, faster floods and it starts to dig down or spread out to make room.
I wanted to see if I could measure that on our own creeks without wading every one of them.
Sources: Paul & Meyer 2001; Walsh et al. 2005
I cut cross-sections across four local streams from USGS 3DEP LiDAR, a laser survey flown from a plane, and compared each channel to what a stable channel that size should look like.
LiDAR can't see under water. Wherever there's water, the lowest point I can measure is the water surface, and every reading says so.
Sources: USGS 3DEP; Nayegandhi, NOAA NGS
Then each reach gets a Rosgen stream type, a method that sorts channels by shape: how entrenched, how wide for its depth, how steep and how winding. Where the data can't support a call, I mark it indeterminate and leave it for a field visit.
Open Four Streams →Sources: Rosgen 1994
After all that counting of what got paved, I wanted to go see what's still here. South Carolina keeps a register of the biggest known tree of every species.
Sources: Clemson Extension registry
Every tree gets points: trunk circumference in inches, plus height in feet, plus a quarter of the crown spread in feet. Biggest score for its species wins.
Try a tree in your yard. Measure around the trunk 4½ feet up.
Sources: American Forests
13 of them are national champions, the biggest of their kind in the country. And 7 are on South Carolina's invasive plant list, like Chinese tallow and tree-of-heaven.
The register measures size. I mark the invasive ones so nobody walks away thinking they're something to plant.
Sources: Clemson Extension registry; SC-IPC 2024
Nobody has walked these with a tape yet. That's the next field day.
Open The Champion Trees →Stuff that's drawn or built but doesn't have a page yet.
A 16 × 40 ft pad fit to 1 m lidar at the west end of the Hunnicutt Food Forest. About 3 cubic yards of cut and 3 of fill, so a wheelbarrow-and-mattock job.
July 2026One printable sheet per walk, a line for every tree, with room for a girth reading and a note.
October 2026