ENEN90028 Chap.6 Spacing, Extent and Support
Spacing, Extent and Support
The scale triplet describes the sampling window. Spacing is the distance or time between observations, extent is the total domain covered, and support is the area, volume or duration represented by one observation. Matching all three to the environmental process prevents rapid variation from being missed, broad trends from being reduced to a patch and important heterogeneity from being smoothed away.
What this chapter covers
- 01
Temporal and spatial spacing
- 02
Overall sampling extent
- 03
Area volume and duration support
- 04
Aliasing patch bias and smoothing
- 05
Process informed scale selection
- 06
Effort trade-offs across and within units
Worked example · free
Repair a dam sampling window
- 1Temporal spacing misses the short pre-dawn oxygen response.
- 1Spatial extent limited to the edge does not represent the dam domain.
- 1Large quadrat support may smooth metre-scale vegetation patches.
- 1Revise the window while protecting independent replication among dams.
Key terms
- Spacing
- The distance or time between adjacent observations.
- Extent
- The total spatial or temporal domain covered by the sample set.
- Support
- The area, volume or duration represented by one observation.
- Aliasing
- Misrepresentation of rapid variation caused by observations spaced too widely.
- Sampling window
- The combined range and resolution through which a process is observed.
Spacing, Extent and Support FAQ
What happens when spacing is too large?
Rapid environmental changes can occur between observations, appear as unexplained noise or disappear entirely. The interval should follow the expected response time rather than convenience alone. Finer spacing is useful only when the added observations still represent the process and the design retains independent units.
How do extent and support differ?
Extent is the total area or time span covered by all observations. Support is the area, volume or duration represented by one observation. Small samples spread widely can have broad extent and small support, while one large composite can have large support but narrow effective coverage.
How should effort be divided?
First protect enough independent sampling units to support the intervention comparison. Then allocate remaining effort to times or sub-samples that adequately characterise each unit. The balance depends on within-unit variability, process scale and the precision required for the endpoint.
Assessment move
Sketch the expected process through space or time, then draw the proposed observations. Label spacing, extent and support with operational units. For every mismatch, name whether the signal is missed, the domain is too narrow or averaging is excessive. Practise trade-off questions that force a choice between more independent units and denser sub-sampling. Build fluency by using the same four-step scale audit on new scenarios.
First state the environmental process and its expected spatial or temporal grain: a pre-dawn oxygen minimum, a storm pulse, a centre-edge gradient or metre-scale vegetation patches. Second write the proposed sampling window with units. Spacing is the interval between observations, extent is the full domain covered by the set, and support is the area, volume or duration represented by one observation.
Third compare each dimension with the process. Wide temporal spacing can alias a brief response; narrow spatial extent can make an edge look like a whole dam; large compositing support can smooth the patchiness that the endpoint is meant to detect. Fourth propose the smallest defensible revision and say what it protects. Practise with paired sketches rather than definitions alone.
Draw a process curve and place observation times on it, then redraw the times around the predicted peak or minimum. Draw a site boundary, mark the covered region and show whether centre and edge are represented. Draw several small quadrats beside one large quadrat and explain how their supports answer different questions.
Always keep the independent sampling unit visible outside these drawings so that denser within-unit measurements are not confused with treatment replication. Use trade-off drills with a fixed field budget. Decide first how many independent sites or treatment units are needed for the environmental comparison. Only then distribute remaining effort among dates, locations and sub-samples needed to characterise each unit.
A proposal that measures one unit exquisitely may still be unable to generalise an intervention effect. For each answer, use the vocabulary precisely: name the scale dimension, identify the signal or domain at risk, and give an operational change such as sampling before dawn, extending transects across the habitat, shortening intervals during a pulse or using smaller distributed supports.
Avoid the vague remedy of collecting more data. More readings at the same unsuitable spacing, within the same restricted extent or with the same excessive support repeat the mismatch. End revision by connecting scale to interpretation. State which endpoint the window represents, which variation remains unobserved and whether the conclusion applies to a moment, a patch, an entire site or the monitoring period.
That closing sentence turns a diagram into a defensible design decision for the team plan, field discussion and report.
