Aerobiology studies biological material in the air: where it comes from, how it travels, how it changes and where it eventually goes. Pollen, fungal spores, microorganisms and fragments of organisms connect living landscapes with atmospheric motion. The gold standard is to follow that whole journey before interpreting a measurement at one place and time.
A pollen count is not a count of nearby trees. A DNA detection is not automatically a living animal at the sampler. A fluorescent particle is not automatically an identified species. Good aerobiology connects the biological source, the transport process and the measurement method, then states exactly which conclusion the evidence supports. This guide explains those connections through original calculations and a fictional monitoring case, rather than treating every airborne signal as the same kind of evidence.
Begin with chapters 1–4 for the field and its mechanisms. Chapters 5–8 explain sampling and concentration; chapters 9–12 examine time, uncertainty and identification. The later chapters connect modern research with source attribution, learning tasks and useful next routes. All numerical examples are invented teaching datasets. They are not measurements of current air conditions, exposure limits or a particular real site.
This subject sits within The Gold Standard and follows the Education parent’s emphasis on learning that transfers. The transferable skill is to turn “something was detected” into a properly bounded explanation of what was measured and what still needs checking.
THE GOLD STANDARD · AEROBIOLOGY
Follow biological material through the air
Choose the part of the journey you need to understand: source, transport, measurement or interpretation.
ROUTE 1 · CHAPTERS 1–4
Follow the airborne journey
Connect biological sources with particle behaviour and transport.
ROUTE 2 · CHAPTERS 5–8
Build the measurement
Separate deposition, collection, air volume and counted fractions.
ROUTE 3 · CHAPTERS 9–12
Interpret the observation
Check averaging, zero counts, identification and biological condition.
ROUTE 4 · CHAPTERS 13–16
Test an explanation
Connect DNA, controls, source attribution and atmospheric budgets.
ROUTE 5 · CHAPTERS 17–20
Preserve and transfer
Check method changes, solve a new case and choose the next route.
Full chapter index · Try the transfer task · Gold Standard subject hub
Chapter index
Follow the airborne journey · Chapters 1–4
Build the measurement · Chapters 5–8
Interpret the observation · Chapters 9–12
Test an explanation · Chapters 13–16
CHAPTER 1 OF 20 · Follow the airborne journey
1. Define the material and the question together
The word biological describes origin; it does not guarantee that an airborne particle is alive. A whole pollen grain, a fungal spore and a fragment carrying DNA are different objects. A useful investigation starts by naming the target and the property of interest. Counting recognisable pollen grains differs from detecting a genetic marker, even if both measurements concern the same landscape.
Aerobiology intersects with several established fields. Palynology follows pollen, spores and related microscopic evidence, including preserved records. Atmospheric science supplies the physical account of moving and mixing air. Aerobiology asks what happens when material from living systems enters that moving medium.
Consider two questions about a park. “Which pollen types were captured this morning?” can be addressed with a suitable collection and identification method. “Which plants grew inside the park?” needs additional evidence about sources and transport. Air may carry material across the boundary. A correct identification at the sampler can coexist with an incorrect inference about where the parent organism lived.
Write the question as an observable target: material, location, time interval and measurement. For example, “the concentration of a defined pollen type at the monitoring inlet during a specified day” is much more precise than “how biological the air was.” The definition guides instrument choice, calculation and interpretation. It also exposes when two apparently conflicting reports actually measured different things.
CHAPTER 2 OF 20 · Follow the airborne journey
2. Follow production, release, transport and arrival
A source must first produce relevant material. That material must then become airborne, remain available during transport and reach the collection point. These are separate stages. Many pollen grains on a plant do not establish a high airborne concentration at a distant station, and a high concentration at the station does not reveal production without considering the intervening processes.
An original bookkeeping example makes the distinction visible. Suppose a fictional source produces 10,000 units during an interval. Assume one-quarter enter the air and one-tenth of those pass through a particular observation region. The corresponding quantities are 2,500 emitted units and 250 arriving units. These fractions are deliberately assigned, not biological constants. If release doubles while production is unchanged, arrival can rise without any increase in the number of source organisms.
The reverse inference is therefore underdetermined. An observed change at the end of a chain could result from a change at several earlier stages. A responsible explanation names the candidate mechanism and the additional observation that could distinguish it. Flowering observations might inform production; wind measurements might inform transport; instrument checks might reveal a measurement change.
This chain also gives a practical reading strategy for research papers. Locate the directly measured stage first. Then identify which other stages were modelled or assumed. A paper measuring captured particles is not necessarily measuring emission rates, even when emissions are the scientific question motivating the work.
CHAPTER 3 OF 20 · Follow the airborne journey
3. Explain why physical size alone is insufficient
Particles differ in shape and density as well as visible dimensions. Aerodynamic diameter expresses behaviour through an equivalent unit-density sphere with the same terminal settling velocity. It therefore need not equal the width seen in a microscope image. The NIOSH chapter on sampling and characterising bioaerosols explains this distinction and its relevance to collection efficiency.
For a learner, the crucial move is to separate a geometrical measurement from a behaviour-based one. Two particles can have similar photographed widths yet interact differently with moving air. Conversely, visually different particles can share similar aerodynamic behaviour. A size label is useful only when its definition accompanies it.
Imagine a report containing two columns both labelled “diameter.” One comes from an image-analysis routine, the other from a device response calibrated against reference particles. Before plotting them on the same axis, ask what each column represents. Agreement is a testable result, not something guaranteed by the shared word.
Avoid drawing a universal distance rule from one particle dimension. Actual travel also depends on the air’s motion, release height, time available and removal processes. A simple settling calculation can clarify one component, but it cannot stand in for the full atmospheric journey. Chapter 4 uses a deliberately simplified model to show both the usefulness and the limitation of such a calculation.
CHAPTER 4 OF 20 · Follow the airborne journey
4. Use a transport calculation without mistaking it for a forecast
Assume a particle starts ten metres above a flat surface and settles downward at a constant 0.02 metre per second relative to still air. Ignore vertical air movement and all other removal processes. The idealised settling time is height divided by speed: 10 ÷ 0.02 = 500 seconds, approximately 8.3 minutes.
Now impose a steady horizontal wind of two metres per second without changing the settling assumptions. During those 500 seconds, horizontal travel is 2 × 500 = 1,000 metres. This is a result of the assigned model. It is not a typical dispersal distance for a named pollen grain or spore.
Changing the release height to twenty metres doubles both the settling time and the horizontal distance in this model. Introducing upward air motion could change the result much more substantially. Turbulence, changing wind and interception by surfaces would require a richer account. The simple calculation is valuable because each dependence is visible, not because the real atmosphere obeys all of its assumptions.
The Atmospheric Science root provides the wider physical setting. When reading a dispersal claim, ask whether the author calculated one idealised path, a distribution of possible paths or an observed arrival. These are different forms of evidence. A precise-looking distance cannot compensate for an unstated assumption about the air carrying the particle.
CHAPTER 5 OF 20 · Build the measurement
5. Choose between concentration and deposition
Concentration describes how much material occupies a volume of air. Deposition describes arrival onto a surface, often expressed per area and time. The two quantities can be connected through a model, but they are not interchangeable. A surface receiving many particles does not by itself reveal how many particles occupied each cubic metre of air above it.
Suppose a fictional collector receives sixty recognisable grains across 0.02 square metre over three hours. The deposition flux estimate is 60 ÷ (0.02 × 3) = 1,000 grains per square metre per hour. No sampled air volume appears in the denominator. Relabelling this answer as grains per cubic metre would be a unit error and an interpretation error.
For comparison, an instrument that draws a measured volume of air through a collection system can support an air-concentration calculation, provided its collection and counting characteristics are addressed. Hirst’s original 1952 volumetric spore-trap paper is a foundational example of this approach: a moving collection surface linked captured material to sampling time.
Begin a measurement plan with the quantity needed for the question. A deposition study may be exactly right for arrival onto a surface. A concentration study may be right for comparing air at an inlet over time. Neither becomes better simply by being expressed with a larger number or a more technical instrument name.
CHAPTER 6 OF 20 · Build the measurement
6. Understand the sampler as a selective measurement system
An instrument does not capture an unfiltered miniature of the entire atmosphere. Material must enter its inlet, be collected or detected, survive relevant handling and be recognised by the analysis. Failure at one stage can change the final result. This is why method details belong with the reported number.
Collection efficiency can depend on particle properties and instrument conditions. The NIOSH bioaerosol methods chapter describes size-dependent collection and explains why an inertial sampler’s stated cut-off is a response characteristic rather than a perfectly sharp boundary. Keep that limited technical point separate from any claim about what a particular environmental sample contains.
Consider a purely mathematical instrument model. If one hundred target particles enter and eighty are recorded, the response fraction is 0.8. Inferring 125 entering particles from one hundred recorded particles assumes that the same fraction applies: 100 ÷ 0.8 = 125. It would be unjustified to borrow 0.8 from a different particle type, flow condition or analysis method without evidence.
There is also a distinction between a transparent correction and a correction that creates false confidence. Reporting a raw observation with a known limitation can be more defensible than applying an unsupported efficiency factor. Record what was corrected, the evidence for the correction and whether uncertainty in that correction is included. A corrected number remains conditional on the model connecting it with the original material.
Our fictional Brookgate monitoring station draws air at twelve litres per minute for thirty minutes. Assume the flow is constant, the timer is correct and all of the relevant collection is examined. The sampled volume is 12 × 30 = 360 litres, or 0.36 cubic metre. If 72 target grains are counted, the uncorrected concentration estimate is 72 ÷ 0.36 = 200 grains per cubic metre.
The unit conversion matters: one cubic metre contains one thousand litres. Dividing 72 by 360 gives 0.2 grain per litre, which is the same concentration as 200 grains per cubic metre. The different numerical values do not describe different air.
| Quantity | Assigned value | Use in the calculation |
|---|---|---|
| Flow | 12 litres per minute | Multiply by elapsed minutes |
| Duration | 30 minutes | Produces 360 litres sampled |
| Volume | 0.36 cubic metre | Denominator for concentration |
| Count | 72 grains | Gives 200 grains per cubic metre |
Now suppose the actual mean flow was ten litres per minute, while the analyst used twelve. The actual sampled volume would be 0.30 cubic metre, giving 240 grains per cubic metre for the same count. An inflated volume estimate produces an understated concentration. This is a useful diagnostic because the error’s direction follows from the denominator. The calculation should be checked against instrument records before a biological explanation is proposed for an apparent change.
CHAPTER 8 OF 20 · Build the measurement
8. Account for a counted fraction without inventing certainty
Sometimes an analysis examines a defined fraction of collected material. Scaling from that fraction to the whole requires a defensible sampling scheme. It does not become valid merely because the fraction can be written as a percentage. A selected patch containing unusually many particles may be a poor representation of the collection.
Extend the Brookgate example with a separate fictional collection. The air volume is again 0.36 cubic metre, but only a representative one-quarter of the collection is examined. Eighteen target grains are observed. The estimated total is 18 ÷ 0.25 = 72 grains, giving the same 200 grains per cubic metre as before.
If an independently validated overall collection efficiency of 0.8 were applicable to this target and configuration, a model-adjusted estimate would be 18 ÷ (0.25 × 0.8 × 0.36) = 250 grains per cubic metre. That extra step is conditional; it is not a routine instruction to increase every count by twenty-five percent.
Keep the measured count, counted fraction, air volume and any response correction as separate fields. This makes the result reproducible and permits later revision if one input changes. It also prevents double correction, such as multiplying a count that was already scaled by the laboratory. A good report can distinguish the instrument-and-analysis estimate from an additional model-adjusted estimate instead of presenting one unexplained final number.
A daily mean can conceal a short peak. An hourly peak cannot be substituted for an entire day. Both may be correct measurements, but they answer different questions about time. A report should state the averaging interval before using words such as “higher,” “lower” or “typical.”
Suppose Brookgate records 100 grains per cubic metre for one hour, followed by 20 grains per cubic metre for three hours. The four-hour time-weighted mean is (100 × 1 + 20 × 3) ÷ 4 = 40 grains per cubic metre. Simply averaging the two reported values gives sixty, incorrectly giving the one-hour and three-hour intervals equal weight.
This matters when data arrive as irregular records. A spreadsheet row is not automatically an equal slice of time. Check interval length, missing periods and whether values are instantaneous estimates or interval means. If a device fails during the highest part of a cycle, the remaining observations may not represent the full period.
For an overall concentration estimated from pooled counts, the corresponding denominator is total valid sampled volume. With constant flow, volume weighting and time weighting coincide. With changing flow, they can differ. State which quantity you intend to estimate before choosing the weights. The arithmetic is not merely a finishing step: it encodes which parts of the observation history are allowed to contribute most strongly to the reported result.
CHAPTER 10 OF 20 · Interpret the observation
10. Treat a zero count as an observation with limits
Zero observed target grains means none were recorded in the examined material under that method. It does not prove that the target was absent everywhere nearby. Limited volume, a short observation interval, incomplete recovery or an identification limit can all matter. The useful next question is what magnitude of concentration the observation could reasonably have missed.
An original probability model illustrates the logic. Assume arrivals are independent, concentration is constant, collection and recognition are perfect, and counts follow a Poisson distribution. If the expected count is three, the chance of observing zero is exp(−3), about 5%. This connects a zero count with the often-used approximate one-sided 95% upper bound of three expected events under this specific model.
For a completely examined air volume of 0.30 cubic metre, three expected events correspond to about ten grains per cubic metre. That is a model-based statistical bound, not a universal instrument detection limit. Clustering, changing concentration or inefficient recovery can invalidate its simple interpretation.
This distinction is useful in learning because it separates uncertainty from ignorance. We may be able to quantify one source of uncertainty while still lacking information about another. Write “zero observed in this volume” first. Add a statistical interval only when its assumptions are appropriate, and do not silently convert it into a statement that a species was absent from the landscape.
CHAPTER 11 OF 20 · Interpret the observation
11. Identify pollen without turning percentages into vegetation cover
Microscopic identification depends on observable features and suitable reference material. A defensible record may resolve a pollen type or broader group without resolving a species. The correct level of identification is the level supported by the evidence, not the most specific name available in a search result.
The MeteoSwiss description of its manual pollen-monitoring method describes microscopic identification and counting followed by concentration calculation. That workflow keeps classification and quantification connected. Counting confidently requires consistent rules for what belongs to each category and how uncertain objects are handled.
Consider two hypothetical air samples. Sample A contains thirty grains of type X and seventy other identified grains; sample B contains thirty X and thirty others. X rises from 30% to 50% of the identified total even though its count is unchanged. If sampled volumes and methods are equal, the example gives no increase in X concentration. The percentage changed because the other component changed.
The specialist teaching guide to palynology develops this compositional reasoning. Aerobiology adds the immediate atmospheric context: release and transport can alter the mixture reaching a station. Neither an airborne percentage nor a sedimentary percentage automatically equals the percentage of surrounding land occupied by the parent plants. Different records require their own evidence about how particles entered the sample.
CHAPTER 12 OF 20 · Interpret the observation
12. Keep detection, identity and biological condition separate
A detected particle, an identified particle and a living organism are three different conclusions. A method may establish one without establishing the others. For instance, a recognisable structural feature can support a morphological classification without demonstrating that the material remains capable of biological activity.
Culture-based results describe what grows under the conditions used; they do not enumerate every biological particle. The NIOSH methods chapter distinguishes viable, culturable and non-culturable material. In this guide, the distinction supports interpretation only; no procedure for culturing unknown airborne material is required for the learning tasks.
Likewise, a fluorescent signal is a measured optical response. In their 2023 Arctic aircraft study, Perring and colleagues explicitly used fluorescent aerosol as a proxy for primary biological aerosol. A proxy is informative because of an evaluated relationship, not because the sensor directly supplies every biological property of interest.
Build a claim ladder in the notebook. First state the measured signal. Next state the identification or proxy interpretation. Finally state any inference about condition, origin or ecological significance. Attach the relevant evidence at each step. If the last step lacks support, stop the conclusion there while retaining the valid earlier observations. This prevents a single instrument output from acquiring several untested meanings as it moves from a technical record into a headline.
CHAPTER 13 OF 20 · Test an explanation
13. Understand airborne DNA as transported evidence
Environmental DNA can reach an air sample in material shed from organisms. Sequence analysis can therefore reveal biological information that a conventional particle count may not provide. But the sample’s location is the location of collection, not automatically the location of the source organism. Transport and analytical identification remain part of the inference.
Tournayre and colleagues’ study published on 2 June 2025 used existing national air-quality monitoring samples for a multi-group biodiversity assessment. It reported more than 1,100 taxa and found that airborne DNA and citizen-science detections supplied complementary information. The authors also treated taxonomic identification cautiously and investigated transport. Their study demonstrates a research capability, not a universal rule that every DNA signal originates within a fixed radius.
For a fictional park survey, suppose a target sequence appears in two samples. Before concluding that the organism lives beside both stations, ask whether the assignments are reliable, controls are satisfactory and the material could have moved between locations. Detection frequency can be a useful measurement without being an animal headcount.
The guide to sedimentary ancient DNA offers a related lesson in a different medium: the route from organism to archive matters. Airborne and sedimentary records are not interchangeable, but both require an account of transport, preservation and contamination before a biological location or history is asserted.
CHAPTER 14 OF 20 · Test an explanation
14. Design controls around the mistake they can detect
A control is useful when it can reveal a particular failure. Naming a sample “control” does not explain what it checks. In an airborne-DNA investigation, for example, handling-related contamination and a genuinely present environmental signal are different possibilities. A blank carried through relevant handling stages can help investigate the former, while repeated environmental sampling addresses a different question.
An original toy dataset illustrates the interpretation. Suppose four environmental samples contain a target signal, but a handling blank also contains it. That does not automatically mean every environmental result is false. It means the simple claim that all detections must have come from the sampled air is no longer justified without investigation. The amount and pattern of signal, method history and independent evidence matter.
The same logic applies outside molecular analysis. An instrument blank may reveal background counts; a co-located comparison may reveal differences between devices; a documented reference material may test classification. None is a substitute for all the others.
Plan the failure question before collecting the control. Write “this check could reveal…” and complete the sentence with a concrete mechanism. Then write what it cannot reveal. A clean handling blank, for instance, does not prove that an environmental sample represents an entire city. It addresses a particular route of error. This discipline makes negative and positive control outcomes informative instead of treating them as ritual boxes on a laboratory form.
CHAPTER 15 OF 20 · Test an explanation
15. Investigate sources with more than a line on a map
A backward trajectory estimates a history of air movement using meteorological information and a model. It can suggest regions worth investigating as sources. It is not a direct observation that a specific particle came from one exact plant or field along the line.
NOAA’s HYSPLIT ensemble tutorial shows how changes in the meteorological information used can produce a spread of trajectories. That spread helps reveal sensitivity. It does not capture every possible source of uncertainty, but it is more informative than presenting one modelled track as an exact biological journey.
Suppose a fictional high-count day has modelled trajectories passing near a wooded region. Three checks would strengthen the proposed source explanation: the target organisms plausibly occur there; their release timing is compatible with the event; and independent observations fit the transport account. The trajectory alone supplies none of those biological facts.
The 2023 Arctic bioaerosol aircraft study combined airborne measurements with trajectory analysis and found higher loadings associated with recent passage through continental boundary-layer air. Its result is bounded by the campaign and methods. It is a useful example of combining observations, rather than proof that every marine bioaerosol signal has a terrestrial origin.
For a source-attribution report, list the alternatives explicitly: local release, transported material, changed mixing and a measurement issue. Ask which observation would make each explanation less plausible. That is a more productive next step than drawing an increasingly detailed map from an unchanged set of data.
CHAPTER 16 OF 20 · Test an explanation
16. Distinguish a changing source from changing atmospheric dilution
A concentration can fall because fewer particles enter a region, because more leave, because removal increases or because the same material becomes distributed through a larger volume. These possibilities are not equivalent ecological stories. The arithmetic should preserve the study boundary before the explanation chooses among them.
Imagine a fixed horizontal area containing 600,000 suspended target particles in a uniformly mixed layer of 3,000 cubic metres. The concentration is 200 particles per cubic metre. If the same number becomes mixed through 6,000 cubic metres, concentration falls to one hundred. No particle destruction or reduction in the source population is needed in that idealised comparison.
Now consider a one-hour budget for another fictional air volume. It begins with one thousand target particles; 300 enter or are released within it; 200 leave by transport; and one hundred are removed to surfaces. The ending stock is 1,000 + 300 − 200 − 100 = 1,000. An unchanged stock does not mean no activity occurred.
Janssen and colleagues’ 2021 fungal-spore study combined observations and global modelling to investigate emission controls. Its differences among emission models and limitations away from source regions illustrate why source estimates depend on the transport and removal account. For the wider budget logic, continue to Atmospheric Chemistry, where stocks, rates and transformations are kept distinct.
CHAPTER 17 OF 20 · Preserve and transfer
17. Preserve comparability when the monitoring method changes
A long time series is not automatically homogeneous. A change of instrument, inlet position, classification system or processing method can create a discontinuity. Before interpreting a jump as environmental change, examine the monitoring history. The metadata may explain a pattern that the plotted values alone cannot.
The MeteoSwiss pollen-station documentation states that automatic measurements began on 1 January 2023. It warns that manual and automatic concentrations may differ and should be used separately until homogeneous series are available. This is a concrete example of an operational authority distinguishing continuity of purpose from direct numerical comparability.
In an invented overlap exercise, an old system reports 40, 80 and 120 units on three days, while a new system reports 50, 100 and 150. The ratios are all 1.25. That small dataset suggests a candidate multiplicative relationship; it does not validate a permanent conversion for all particle types, seasons and concentration ranges. The next task is to test the relationship on independent conditions.
A practical continuity record
Keep the instrument identifier, location, valid sampling periods, target definition, analysis version and known changes beside the measurements. Preserve original values when producing adjusted series. A future reader should be able to separate what the instrument reported from what a harmonisation model inferred. Otherwise, later improvements can erase the evidence needed to understand why the historical record changed.
CHAPTER 18 OF 20 · Preserve and transfer
18. Transfer task: repair a false increase
Your task
Two fictional reports from Brookgate are placed side by side. Report A records sixty grains from 0.30 cubic metre of air and examines the complete collection. Report B records ninety grains from 0.60 cubic metre and also examines the complete collection. The analyst says: “The count rose by half, so the concentration rose by half.” Assume comparable methods and no response correction.
Calculate both concentrations and the percentage change from A to B. Then repeat B’s calculation under a changed condition: the ninety grains came from a representative half of the collection, not all of it. Explain which single piece of metadata changes the interpretation.
Finally, an afternoon interval contains a higher percentage of type X pollen than a morning interval, but X’s concentration is unchanged. Give one arithmetic explanation that does not require increased X production.
Answer and reasoning
A’s concentration is 60 ÷ 0.30 = 200 grains per cubic metre. B’s is 90 ÷ 0.60 = 150 grains per cubic metre. The change is (150 − 200) ÷ 200 × 100 = −25%. More grains were counted, but more air was sampled; concentration fell by one-quarter.
Under the changed counting condition, B’s estimated total becomes 90 ÷ 0.5 = 180 grains. Dividing by 0.60 cubic metre gives 300 grains per cubic metre, a 50% increase over A. The counted fraction changes the answer. Without it, the original record was incomplete.
Type X’s percentage can rise if other pollen types decline while X remains unchanged. The lesson is to reconstruct the measurement before explaining the biology. Counts, fractions, air volume and category totals all belong to the evidence, even when a chart displays only the final percentage.
CHAPTER 19 OF 20 · Preserve and transfer
19. Questions that make an aerobiology report clearer
Does clear air contain no biological material?
Visible clarity is not a particle census. A statement about a target requires a method capable of detecting it. “I could not see anything” describes an observation at a particular visual scale, not a measurement of every microscopic component.
Does a high pollen count identify the nearest plant?
No. Identification of the captured material and identification of its source location are separate tasks. Nearby vegetation may be relevant, but transport, release timing and the monitoring position must also be considered.
Can air-quality mass measurements replace pollen counts?
They measure different quantities. Mass concentration does not by itself provide the number or identity of pollen grains. A conversion would need information about the particles and their composition; it cannot be assumed from shared use of “per cubic metre.”
Is a DNA result a species abundance estimate?
Not automatically. Detecting a sequence and estimating organism abundance require different measurement models. Sequence assignment, shedding, transport, recovery and analytical response can all stand between organisms and reported detections. State the supported detection result before attempting a stronger quantitative inference.
What can students investigate without specialised collection?
Use documented public monitoring data or supplied teaching images. Check units, valid intervals and method history; calculate a time-weighted mean; or compare two interpretations of the same observations. These tasks develop real scientific reasoning without requiring unknown airborne material to be cultured.
What makes a conclusion dependable?
It names the target, method, place, interval and main limitation. “This method recorded a higher concentration of this pollen type during this interval” is more useful than an unsupported statement that an entire ecosystem became more biologically active.
CHAPTER 20 OF 20 · Preserve and transfer
20. Build an investigation and choose the next specialist route
Start with one question whose answer would change your understanding. Define its target and sampling interval, then identify the measurement capable of addressing it. Before analysing any values, inspect the metadata. A beautiful time-series plot cannot repair an unknown air volume, a changed location or a missing definition of the counted material.
Build the explanation in four passes. First describe the observation without naming a cause. Next calculate the relevant quantity with units. Then propose a mechanism connecting source, transport and measurement. Finally identify a competing explanation and the evidence that would help distinguish it. A strong investigation may end with a narrower conclusion than expected while still producing useful knowledge.
For further learning, follow the part of the chain that remains uncertain:
- Pollen identity and environmental archives: The Gold Standard of Palynology and the specialist pollen-records teaching guide.
- Motion, mixing and atmospheric boundaries: Atmospheric Science and Atmospheric Chemistry.
- Molecular traces and archive interpretation: sedimentary ancient DNA analysis and Tournayre et al.’s 2025 airborne-DNA research.
- Measurement history and current data definitions: MeteoSwiss pollen-station documentation.
- Source and transport models: Janssen et al., 2021 and NOAA’s trajectory-ensemble teaching example.
Sources were checked on 23 October 2026. Historical methods papers and dated research are identified as such. For a wider process question, continue through How X Works; for another subject, return to The Gold Standard hub. Carry the same discipline into the next field: define the material, follow its route and keep the conclusion within the evidence.