Field protocol

Measure the street,
not the satellite.

Paired air readings, recorded at walking height. The same field protocol across six routes in Merewick.

The question

What changes
under a canopy?

We compared air temperature at nearby shaded and open-sky points, while keeping the route, time window and sensor height consistent.

We measured air temperature, not surface temperature. The two answer different questions.

Diagram of paired shaded and open-sky observation points along a walking route — Route diagram. Nearby pairs reduce differences in weather exposure; they do not remove differences in buildings, wind or paving.

Repeatable steps

One protocol, every stop.

Choose the pair

Select one shaded and one open point within 30 metres. Record paving, building height and visible canopy cover.

Let the sensor settle

Place a shielded air sensor at 1.5 metres. Wait five minutes before logging. Avoid contact with walls, bodies and hot surfaces.

Repeat the circuit

Visit every pair at 10:00, 13:00 and 16:00. Alternate the route direction between days to reduce order effects.

Keep the pair together

Calculate sun-minus-shade differences for each pair, then summarise by time window. Never treat repeated readings as independent sites.

Section drawing of a tree canopy and two shielded air sensors above the pavement — Measurement geometry: shaded and open-sky sensors at the same height. The tree drawing is schematic, not a surveyed specimen.

Instruments & timing

Small details.
Comparable readings.

Instrument: Shielded digital air thermometer
Stated accuracy: ±0.3°C
Field dates: 14–16 July 2026
Observation windows: 10:00 / 13:00 / 16:00
Canopy estimate: Upward photo, manual grid count

From notebook to figure

Keep the raw context.

Each observation retained its route, time, sky condition and surface notes. Incomplete pairs were excluded from paired differences. No missing readings were imputed.

Inspect the sample dataset ↗

The evidence, in view

Look at the figures.

Read the observations alongside the method, uncertainty and limits.

Read the findings