The Gold Standard · Ecology
Make the decision path visible
When a monitoring site returns after an outage, the approval queue can fill quickly. Backfilled records, new calibrations, storm events and the next seasonal field window may all compete for the same reviewers. First-in, first-out feels neutral, but it can leave a high-consequence data defect unresolved while routine records move ahead.
The gold standard triages by scientific and operational risk, protects fixed seasonal decisions, keeps provisional status visible and never turns queue pressure into silent approval.
Choose the route closest to your decision
Chapter contents
Chapters 1–3
Chapters 4–6
Chapters 7–9
Chapters 10–12
CHAPTER 1 OF 12 · MAP THE QUEUE
1. Define the unit waiting for approval
A queue item may be a day, event, sensor interval, calibration period or complete water year. Choose units that match the review decision. Oversized items hide partial progress; tiny items create administrative noise.
Record site, parameter, date range, method era, provisional status, dependencies and named owner. A queue that exists only in email cannot support reliable triage.
CHAPTER 2 OF 12 · MAP THE QUEUE
2. Name the deadlines and why they matter
Seasonal deadlines are tied to ecological decisions: migration surveys, spawning windows, low-flow alerts, restoration actions or public reporting. Distinguish a fixed decision date from an internal target.
Late approval can be harmful, but premature approval can be worse. State what can proceed with provisional data, what needs an interim caveat and what must wait for a completed review.
Practical check
Write the decision, evidence and limit in three separate sentences. If one sentence depends on an unstated assumption, name it before continuing.
CHAPTER 3 OF 12 · MAP THE QUEUE
3. Describe risk in both directions
Error risk includes releasing a wrong value, missing a real event or joining incompatible calibration eras. Delay risk includes losing a field window, withholding a safety signal or making a management choice with stale evidence.
Score consequence, likelihood, detectability, reversibility and time sensitivity. Do not use ecological importance alone; a highly important series may still have a simple, well-controlled review.
| Field | Why it matters |
|---|---|
| Approval unit | Defines the decision |
| Dependencies | Prevents premature review |
| Deadline | Connects to ecology |
| Owner | Prevents orphan work |
CHAPTER 4 OF 12 · TRIAGE OPENLY
4. Create an explicit triage rule
Build priority bands with entry criteria. Critical items may include high-consequence anomalies, records needed before a fixed seasonal action or changes that affect many downstream products. Routine stable intervals can follow a standard lane.
Publish the rule to analysts and reviewers. Exceptions need a named decision-maker and a reason. Otherwise the queue will be reordered by status, persistence or whoever asks most loudly.
CHAPTER 5 OF 12 · TRIAGE OPENLY
5. Protect prerequisites and linked records
Approval often depends on calibration, datum checks, discharge ratings, laboratory results or neighbouring sites. Map these dependencies so a downstream item is not reviewed repeatedly while its prerequisite remains unresolved.
Bundle related evidence where it helps consistency, but do not let one difficult record block every unaffected period. Split the item and mark the boundary.
Practical check
Write the decision, evidence and limit in three separate sentences. If one sentence depends on an unstated assumption, name it before continuing.
CHAPTER 6 OF 12 · TRIAGE OPENLY
6. Worked example: spring migration meets an aging backlog
A recommissioned wetland station has 90 days of routine data, two storm events and a calibration change. The spring migration assessment is due in 14 days. Review capacity is 40 hours: routine data need 18, storm events 16 and the calibration bridge 12.
The calibration bridge is reviewed first because all later values depend on it. One storm event needed for the migration model follows; the second storm and routine intervals remain provisional with clear status. Changed-context check: if the unreviewed storm triggered a public-health threshold, consequence and urgency would move it ahead even if it is not needed for migration.
| Dimension | Question |
|---|---|
| Consequence | What if wrong? |
| Urgency | What closes with time? |
| Detectability | Would error be noticed? |
| Reversibility | Can action be repaired? |
CHAPTER 7 OF 12 · REVIEW PROPORTIONATELY
7. Match review depth to risk
Use a documented standard review for stable, low-risk intervals and deeper independent review for calibration changes, improbable extremes, method transitions or values that drive irreversible action.
Risk-based does not mean evidence-light. Every lane retains minimum checks, provenance and approval authority. Higher risk adds depth; lower risk does not remove accountability.
CHAPTER 8 OF 12 · REVIEW PROPORTIONATELY
8. Keep provisional data visibly provisional
Users should see approval status in downloads, dashboards and citations. Explain that provisional data may change and identify the date of the latest review.
Do not copy provisional values into a product that loses the status flag. If provisional use is allowed, record the decision, sensitivity to revision and a follow-up owner.
Practical check
Write the decision, evidence and limit in three separate sentences. If one sentence depends on an unstated assumption, name it before continuing.
CHAPTER 9 OF 12 · REVIEW PROPORTIONATELY
9. Use independent review where it changes trust
Independent review is most valuable when judgement is consequential or the analyst built the correction being assessed. Specify the questions the reviewer must answer rather than requesting a vague sign-off.
For scarce specialists, use targeted review of the bridge, anomaly or model while standard checks proceed elsewhere. Independence should focus attention, not duplicate every keystroke.
| Lane | Typical item |
|---|---|
| Critical | High-consequence anomaly |
| Seasonal | Fixed decision window |
| Dependent | Bridge or calibration prerequisite |
| Routine | Stable controlled interval |
CHAPTER 10 OF 12 · CLOSE AND LEARN
10. Control queue aging and escalation
Track age by priority band, blocked reason, reviewer and promised date. Escalate when a seasonal deadline approaches, a critical item lacks an owner or provisional use is expanding.
Aging targets should prompt action, not automatic approval. Options include adding qualified capacity, narrowing a release, issuing an interim product or deferring a dependent decision.
CHAPTER 11 OF 12 · CLOSE AND LEARN
11. Close the item with a release record
Approval records should identify reviewed range, evidence, corrections, reviewer, effective date and downstream products notified. Preserve prior provisional values where audit or citation history requires them.
If a later audit revises approved data, link the revision to the original release and explain the impact. Approval is a controlled state, not a claim that future correction is impossible.
Practical check
Write the decision, evidence and limit in three separate sentences. If one sentence depends on an unstated assumption, name it before continuing.
CHAPTER 12 OF 12 · CLOSE AND LEARN
12. Improve the system after each seasonal cycle
Compare planned and actual review hours, bottlenecks, exception reasons and late decisions. Repeated emergencies may indicate poor capacity planning, unstable instrumentation or review units that are too large.
The gold standard keeps scientific standards fixed while changing the workflow around them: earlier calibration checks, cross-trained reviewers, better status metadata and decision-specific release plans.
| Check | Record |
|---|---|
| Status | Approved or still provisional |
| Scope | Dates and parameters |
| Impact | Products and decisions |
| Lineage | Reviewer and version |
Frequently asked questions
Is first-in, first-out ever appropriate?
Yes for comparable routine items, but it should not override a fixed ecological deadline or a high-consequence dependency.
Can provisional data be used in a seasonal decision?
Sometimes, if the status, uncertainty, sensitivity and follow-up are explicit. High-consequence irreversible actions may require approval first.
Does risk-based review lower the standard?
No. It preserves minimum checks for every item and adds depth where error or delay has greater consequence.
What should happen when the queue exceeds capacity?
Narrow the release, add qualified review, protect dependencies and disclose delays. Do not silently approve or remove status labels.
Continue through verified routes
Core routes
Ecology subject root · Gold Standard of Education · Gold Standard directory · How X Works hub
Specialist routes
- Recommissioned-site data approval and provisional-record aging
- Recommissioned-site calibration transfer and seasonal drift
- Recommissioned-site drift surveillance and second-break detection
- Sentinel-network recommissioning and reconnecting broken time series