← Services
Service · 02

Climate fidelity protocol for controlled environment units

Short, targeted protocols that measure how faithfully a controlled-environment unit reproduces a prescribed climate, and where its operational limits actually lie.

What we do

Existing chamber standards cover static setpoints and, at best, periodic cycles for a single variable. Modern controlled environment units, and ecotrons in particular, increasingly operate in a regime that has no shared benchmark: dynamic, multivariate, physically coherent forcings. The climate fidelity protocol fills that gap.

We build short, one- to two-week test forcings that sample representative time-series across the full distribution of conditions the unit is expected to reproduce, with a gradual ramp-up of extremes. Instead of stumbling into the operational envelope through months of trial and error, the unit is walked through it in a controlled, informative order, so its real limits and behaviour are known quickly and precisely.

We treat external perturbations as first-class inputs: outside radiation and cloud, boundary conditions, seasonal context. Deviations between prescribed and realised time-series are then attributed to identifiable causes rather than lumped as noise, and the resulting recommendations act on the right lever: control tuning, calibration, or hardware.

Deliverables
  • 01Per-unit capability datasheet, scored against every prescribed metric
  • 02Narrative technical report with attributed prescribed-vs-realised gaps
  • 03Recommendations for calibration, control tuning and hardware refinements
  • 04Optional retrospective fidelity analysis from existing facility logs
Who it is for

Four situations where fidelity testing pays off quickly.

01
New units in commissioning

A rigorous baseline before the first campaign, so the unit's real envelope is known and documented from day one.

02
Engineering teams designing custom units

A transparent, shared performance target to design, build and iterate against, well beyond static setpoint checks.

03
Facilities after upgrades or retrofits

A short, smart verification campaign that shortens time-to-conclusion compared to ad-hoc trial and error.

04
Facilities with long prescribed-vs-realised histories

A retrospective fidelity analysis on your existing logs, with no test time required and the same reporting depth.

Forcing classes

Four classes of forcing, from steady state to fully coherent.

Not every unit needs to reproduce every class of forcing, and not every class stresses a unit in the same way. A fidelity protocol picks the classes that matter for your science and evaluates each with metrics designed for it.

FC-1
Static

Constant setpoints held over a block. The baseline: does the unit hold?

FC-2
Periodic

A deterministic, repeating cycle, e.g. a diurnal temperature wave.

FC-3
Aperiodic, per variable

Non-repeating prescribed series; each variable evaluated on its own.

FC-4
Aperiodic, multivariate coherent

As FC-3, but variables share one physically coherent source, with covariance preserved end-to-end.

Metrics by class

What "fidelity" actually measures.

Each class inherits the previous one's metrics and adds those it requires. The result is a compact, comparable scorecard that captures both first-order accuracy and higher-order dynamical behaviour.

FC-1 · Static
  • 01Bias
  • 02Drift
  • 03Noise
FC-2 · Periodic
  • 01Bias
  • 02Amplitude ratio
  • 03Phase preservation
  • 04RMSE vs. cycle
FC-3 · Aperiodic, per variable
+ FC-2 metrics
  • 01Multi-lag autocorrelation
  • 02ETCCDI-style extrema
  • 03Response latency
FC-4 · Aperiodic, multivariate coherent
+ FC-3 metrics
  • 01Rank correlation
  • 02Wasserstein distance
  • 03Copula fits
  • 04Event co-variance
Workflow

From unit under test to capability datasheet

  1. 01
    01
    Scope & forcing class

    We agree on the target climate, the forcing classes to test (FC-1 through FC-4), the variables that matter for your science, and the operating envelope to explore, including where extremes should be probed.

  2. 02
    02
    Test forcing generation

    We build a one- to two-week prescribed forcing that samples representatively across the distribution, with a gradual ramp-up of extremes so the unit's limits are approached in a controlled way.

  3. 03
    03
    Run, measure, contextualise

    The protocol runs on the unit while we log realised conditions alongside external perturbations (outside radiation and cloud, boundary conditions) so that any gap between prescribed and realised can be attributed rather than lumped as noise.

  4. 04
    04
    Score, report & recommend

    Every prescribed metric is computed per forcing class and delivered as a per-unit capability datasheet, with a narrative report and concrete recommendations for calibration, control tuning or hardware refinements.

Two service modes

A live campaign, or a retrospective analysis.

Mode A
Live capability test

A short one- to two-week campaign on the unit itself. Best for new units, custom builds, and post-upgrade verification. We supply the test forcing, we score the realised time-series, you receive the capability datasheet and report.

Mode B
Retrospective fidelity analysis

For facilities with a long history of prescribed-vs-realised logs, we analyse the existing record (no test time required) to characterise the unit's fidelity, identify limitations and propose targeted improvements.

Start a project

Know your unit's real envelope.

Tell us about the unit, the science it supports, and whether you're commissioning, retrofitting, or auditing an existing record. We'll propose the right fidelity protocol.

Contact us →