Knowledge is the asset this project is built on — see Vision. This directory is where it accumulates.
The documents that run it
| Document | Answers |
|---|---|
| Domain map | What we need to know — four layers, ~40 sub-domains, and how they are prioritised |
| Research protocol | How we find it out — the six-step sprint, from key questions to open questions |
| Source registry | Where it comes from — the A–E evidence tiers and a starting directory of sources |
| Wave 0: actor mapping | Who is already doing this — the ecosystem programme that runs first |
| Watchlist | How we stay current — the standing monitoring system Wave 0 leaves behind |
The evidence protocol covers the other half: how we prove that an intervention actually worked on a real site.
Artefact types
Six kinds of document, five with templates. Everything we learn lands in one of them.
| Artefact | Purpose | Lives in | Template |
|---|---|---|---|
| Domain brief | The living synthesis for one sub-domain: what we know / assume / don’t know | domains/ | template |
| Source note | One per source: extracted numbers, method, limits, bias | sources/ | template |
| Solution card | One per candidate intervention: mechanism, effect, cost, preconditions, measurement plan | solutions/ | template |
| Metric definition | One per measured quantity: unit, instrument, method, baseline needs, benchmarks | metrics/ | template |
| Actor profile | One per ecosystem actor: what they do, publish, claim, and who they reference | actors/ | template |
Live content: domains/ · solutions/ · sources/ ·
actors/
| Case study | One per completed intervention: baseline → prediction → result | cases/ | — (follows the evidence protocol stages) |
Directories are created when there is something to put in them, not before.
The shape of the whole system
domain map ──▶ key questions ──▶ source sweep ──▶ source notes
│
▼
open questions ◀── CHALLENGE ── domain brief
│ │
┌────────────────┼───────────────┐ ▼
▼ ▼ ▼ solution cards
measurement next sprint content │
backlog backlog ▼
│ metric definitions
└──────────────▶ Site Zero / client site ◀────┘
│
▼
evidence protocol: baseline → prediction
→ intervention → verification
│
▼
case study ──▶ back into the domain brief
as Tier A evidenceThe loop closes: research generates things to measure, measurement generates evidence, evidence upgrades the research from assumed to known.
Conventions
- Tier every claim. A, B, C, D or E — defined in the source registry. Untiered numbers are a defect.
- Date every fact, including the date of the underlying data, not just the publication date.
- Units always. Quantity, unit, and the conditions under which it was measured.
- Ranges over point estimates wherever the underlying reality is a range.
- “What we don’t know” is a required section, not an optional one.
- Naming: domain files
<ID>-<slug>.md(e.g.W7-pool-water-demand.md); source notes<year>-<slug>.md.
Status
Wave 0 rounds 1–2 are done. 46 of 57 seed candidates verified, 17 actors found by snowball, 11 still unchecked. Four sources are verified against publisher metadata, two domain briefs exist (W7 pool water, C5 permitting) and the first solution card is written.
No full text has been read yet. W7 has eight Tier B claims verified at abstract level; C5 is entirely Tier D and carries a warning not to advise anyone on it.
Next: read BOIB primary text on the rustic-land legalisation window (the only finding with a deadline attached), contact the UIB water research group, and read the two Balearic pool papers in full.
In parallel, starting immediately: instrument Site Zero. The baseline clock is the one thing that cannot be caught up on later.
Then Wave 1, per the domain map: measurement and instrumentation (P1), the pool (A1), water sources and losses (W1–W4), context, and energy demand disaggregation (E4).