Haute Lumière
Commerce · VII.06 · MMXXVI · daylight
For the person with a notebook, a term, and an internet connection. Everything in this workbook can be done for nothing, from anywhere, and most of it will tell you something about where you live that nobody has told you before.
Most environmental exercises ask you to estimate something you cannot check. This one does the opposite. You are going to close a set of books.
By the end of the term you will have produced a one-page water account for a real catchment, with a residual you cannot explain away, and you will have priced a real payment scheme with all three of its terms. Nobody will have handed you the numbers. You will have found them, and you will know exactly which ones you had to assume.
That last part is the skill. An economist who can say "here are my nine inputs, three of them are measurements and six are assumptions, and here is what the answer does when I move the six" is doing something a great many professionals cannot do. It is not hard. It is just rarely asked for.
Exercise 1.1 — Locate yourself in a catchment (one afternoon)
Find out which watershed you are standing in. Every country with a hydrological service publishes catchment boundaries, and most publish them as maps you can click. Write down, in your own notebook:
Then walk to the nearest watercourse and stand in it if you can. This is not a sentimental instruction. The rest of the term is arithmetic, and the arithmetic is about something. Go and see the something once.
Exercise 1.2 — Find the gauge (one week)
Somewhere in your catchment, or immediately below it, there is almost certainly a stream gauge, and its record is almost certainly public. Find it. Record:
| Station name and number | |
| Who operates it | |
| First year of record | |
| Measurement interval | |
| Where the data is published | |
| What it costs to download |
If your catchment has no gauge, that is a finding and you should write it down as one. Note where the nearest one is and what it would take to put one in.
Exercise 1.3 — The appreciative interview (one conversation)
Find one person who has worked on water where you live — a farmer, an angler, a utility engineer, a council officer, a hydrologist, someone from a river trust — and ask them three questions. Not what is wrong with the river. These:
When has water been managed well here? What made that possible? What would it take to have more of that?
Write down their answers verbatim before you interpret them. You are collecting the positive core, and the positive core is almost never in a report.
Exercise 2.1 — The basin identity, on your basin (two weeks)
P = ET + Q + ΔS + ε
Fill in the table for one recent water year. Mark each entry M for measured or A for assumed. Be strict about this; it is the whole exercise.
| Term | Value (mm/yr) | Value (Mm³/yr) | M or A | Source |
|---|---|---|---|---|
P precipitation | ||||
ET evapotranspiration | ||||
Q streamflow at the gauge | ||||
ΔS change in storage | ||||
ε the residual | computed |
For ET, do not guess. Use Zhang, Dawes and Walker (2001):
ET / P = (1 + w·(E₀/P)) / (1 + w·(E₀/P) + (P/E₀))
w = 2.0 for forest, w = 0.5 for grass
E₀ = potential evapotranspiration
Run it for your basin's dominant cover. Worked, for the Catskill/Delaware basin at P = 1,200 mm and E₀ = 700 mm: forest gives ET = 670 mm and Q = 530 mm; pasture gives ET = 516 mm and Q = 684 mm.
Exercise 2.2 — Interrogate your own residual (one week)
Whatever ε came out as, write 300 words on what it could be. Three candidates and no fewer: an unrecorded abstraction, a gauge problem, a storage estimate problem. Say which you think is most likely and what one piece of evidence would settle it.
If your residual came out as zero, you have almost certainly assumed one of the terms from the others. Find which. That is the most valuable mistake in this workbook and nearly everyone makes it once.
Exercise 2.3 — The land-cover trade (one week)
Run Zhang twice for your basin — once at w = 2.0, once at w = 0.5 — and compute the yield difference. Then convert it:
yield difference (mm/yr) ÷ 1000 × basin area (m²) = m³/yr
Worked, for the Catskill/Delaware basin: the difference is 154.3 mm a year, which across 4,047 square kilometres is 451.9 million gallons a day — against a city draw of 900 mgd. Keeping that basin forested costs New York about half as much water as New York drinks.
Now the question, and answer it in writing: in your basin, is that trade a good one? It depends entirely on whether your basin is short of quantity or short of quality, and you now have the arithmetic to say which.
Exercise 2.4 — Sensitivity, done properly (one week)
Take the three inputs you marked A and vary each by ±20 percent. Produce a small table showing what your answer does. Then write one sentence: "the conclusion survives / does not survive the assumptions, and here is the one input it is most sensitive to."
This single habit will distinguish your work for the rest of your career.
Exercise 3.1 — The downstream price (one week)
Find the water utility that serves your area. Its published accounts or regulatory submissions will tell you its capital programme. Find one treatment or storage project it is planning, and record:
Then compute the realised unit cost from the completed project and re-price the planned one. Worked, from the chapter: New York's Croton plant came in at $3.2 billion against an original $800 million — a 4.0× overrun — for 320 million gallons a day, which is $10.0 million of capital per mgd.
Annualise it with the capital recovery factor:
CRF = i / (1 − (1 + i)⁻ⁿ)
At 4 percent over 40 years, CRF = 0.050523, so $9.0 billion carries $454.7 million a year.
Exercise 3.2 — All three terms (one week)
A* = t / (b − c)
Estimate each for your basin, and mark each M or A. Then compute A* and go and find out the median holding size in your catchment — agricultural censuses publish it.
| Your basin | Worked: median basin | |
|---|---|---|
b downstream benefit / ha / yr | $115.71 | |
c upstream opportunity cost / ha / yr | $80 | |
t transaction cost / contract / yr | $1,200 | |
A* | 33.6 ha | |
| Median holding | 2 ha | |
| Verdict | short by 17× |
Exercise 3.3 — Design the aggregation (one week)
If your A* is larger than your median holding, the scheme cannot work as written — and the fix is structural, not scientific. Identify one real organisation in your catchment that could hold a single contract on behalf of many landholders. A farmers' union, a co-operative, a commons association, a river trust, a parish council.
Write one page on what it would need to be able to say yes: governance, staffing, a disclosed administration margin, and who it answers to. Worked: one 3,000-hectare contract carries 40 cents a hectare a year of transaction cost and clears by 89×, replacing 1,500 contracts and 1,124,250 pairwise relationships with one.
Exercise 4.1 — The four failure modes, checked against your basin
For each, write two sentences: could it happen here, and what would be the first visible sign?
Exercise 4.2 — The sentence that protects the scheme
Write, in one sentence, the clause you would put into the instrument on day one so that the year-four audit does not kill it. A model, which you should improve on: this is a purchase of a deferred capital decision, settled at the intake against the regulatory threshold, and it was never a purchase of a per-parcel hydrological outcome.
Exercise 4.3 — Go back to the water
Return to the watercourse you stood in during week one. Take your account with you. Read the residual out loud, standing there.
This is the only exercise in the workbook with no deliverable, and you should do it anyway. The pleasure of a set of books that close is real, and it is worth knowing what it feels like in your body rather than only on a page.
Produce a water account and a scheme appraisal for a single real catchment.
Section 1 — The basin (2 pages). Boundary, area, gauge, land cover, who draws from it and how much. One map you made or annotated yourself.
Section 2 — The account (3 pages). The identity, every term, every source, every assumption marked. The residual, and your three candidate explanations for it. A sensitivity table.
Section 3 — The alternative (2 pages). The engineered option your basin's utility would otherwise build, priced from a completed project rather than an estimate, annualised with the capital recovery factor.
Section 4 — The three terms (2 pages). b, c, t, and A*. The median holding. The verdict, stated as a number and not as an adjective.
Section 5 — The design (2 pages). The aggregation, the settlement point, the protective clause, and the walk-away price. Worked, from the chapter: deferring a $9.0 billion plant by ten years at 4 percent is worth $2.92 billion, against a programme costing $450.6 million in present value — a net present value of $2.47 billion and a ratio of 6.5 : 1, with a walk-away price of $360 million a year.
Section 6 — What you could not find out (1 page). Named, listed, and honest. This section is marked as heavily as any other. A survey that hides its gaps is an advertisement.
Score yourself honestly. The scale is the work itself, not other students.
| Not yet | Partly | Yes | |
|---|---|---|---|
| I can write the basin identity from memory and say why it closes | |||
| I marked every input M or A and can defend each mark | |||
| My residual is a real number I did not engineer to zero | |||
I computed ET from a published relation rather than assuming it | |||
| I priced the alternative from a completed project, not an estimate | |||
I printed all three terms of the scheme, including t | |||
I know my basin's median holding and what it does to A* | |||
| I can say what my conclusion is most sensitive to | |||
| I named what I could not find out | |||
| I can explain, in one sentence, why settlement happens at the intake |
Three things worth taking out of this term into work that has nothing to do with water.
Ask what the boundary is before you believe the number. Most figures you will meet professionally have no set of things they must add up to. Notice which ones do. Those are the ones you can audit.
Print the third term. In any transaction between two parties, someone is paying the cost of making the transaction happen, and it is almost never on the page. Find it and put it there. It changes decisions.
Move the settlement point to where measurement already happens. When causation cannot be demonstrated, do not commission a study to demonstrate it. Find the measurement somebody is already compelled to make and settle against that instead. It is cheaper, it is faster, and it survives the audit.
Three per D, to be asked out loud with other people in the room.
Discovery. When has water been managed well where any of us grew up — who made that happen? · What does someone upstream of us already do, unpaid, that we benefit from? · Which measurement in our basin does everybody already trust, and what made it trustworthy?
Dream. If our catchment published audited accounts, what would we learn first? · If entitlements here were shares rather than volumes, which argument would simply stop? · Who would become interesting who is currently ignored?
Design. Which single organisation here could hold one contract for everybody upstream? · What is already measured for a regulator that could serve as a settlement point for free? · If we could buy only one upstream practice, which one, and how would we know we had it?
Destiny. What would keep this account being published when none of us are here? · Whose balance sheet would have to move? · If the payment outlived the practice, who would notice first, and what could they do about it?