How to read the scientific record

Start with the measurement

Every measurement has a meaning, a unit and a scope. An average over the published world cells describes something different from a local reading or an individual state. Population totals can change when membership changes, as well as when individuals accumulate experience.

Simulation time describes elapsed time within the model. It is separate from the calendar date of a Journal article. A dated update must not be mistaken for a timestamp inside the simulated world.

Distinguish stocks, counters and estimates

Fruit dry matter is a stock available at an instant. Produced dry matter is cumulative since ecology initialization. A cumulative total is not a rate, and ripe fruit is already part of the total fruit stock. Living model plant entities, cover-linked resource points and species references are also different counting units.

Shared learning transitions and bodily outcomes describe different aspects of activity. A transition records experience accepted by the common learning system; it is not a policy-update batch, a skill or proof of behavioural benefit. Earlier individual-learning counters belong to their original model and cannot be treated as a continuation of this shared series.

Acoustic power, metabolic power and received sound intensity are different physical quantities. Frequency is unavailable without a qualifying sound signal. A numerical zero is meaningful only when that quantity was actually measured as zero.

Keep a stable beginning

Time zero means the state before the first step, not a table filled with zeros. Water, biomass and body state may already have non-zero values. A later snapshot cannot be used to reconstruct an unrecorded beginning.

Graphs connect retained observations rather than reconstructing everything between them. Missing readings leave gaps. Display-unit preferences change the presentation; CSV downloads retain source units and missing-value status. A change in world extent also changes what an environmental total covers.

Keep the limits visible

A missing measurement is not zero. A snapshot describes a state at one moment; it does not establish what happened before that moment. Historical trends require a suitable sequence of measurements.

Aionterra is a model, not a direct measurement of the real world. Results depend on assumptions, inputs, resolution and numerical methods. Interpretation should remain within what the available record can support.