Biometeorology studies the exchanges between living organisms and the atmosphere: how weather affects a leaf, an animal or a person, and how living systems alter their surroundings. Its gold standard is to connect the conditions an organism actually experiences with a defensible explanation of its response. A regional temperature reading is a starting point; the useful question is what reaches the organism, for how long, and through which physical and biological processes.
This guide follows heat, water, seasonal timing and observation through a fictional garden beside a paved courtyard. Two plants share the same weather forecast but experience different radiation, airflow and water supply. Their responses help us understand why the same weather can produce different biological outcomes, and how to investigate those differences without confusing a plausible mechanism with a demonstrated cause.
All numerical datasets and intervention results below are invented teaching examples. They illustrate reasoning, not current conditions, species-specific thresholds or instructions for managing heat exposure. The guide belongs to The Gold Standard and follows the Education parent’s emphasis on learning that transfers: understand a relationship, test its assumptions and use it when the setting changes.
THE GOLD STANDARD · BIOMETEOROLOGY
Understand the weather an organism experiences
Choose a route through exposure, heat, water, seasonal timing and evidence. Each chapter stays open and links back to the index.
ROUTE 1 · CHAPTERS 1–4
Connect weather with exposure
Define the organism, microclimate, energy balance and humidity.
ROUTE 2 · CHAPTERS 5–8
Follow exchange and timing
Connect water transfer, budgets and seasonal development.
ROUTE 3 · CHAPTERS 9–12
Read the living response
Separate detection, behaviour, human conditions and canopy feedback.
ROUTE 4 · CHAPTERS 13–16
Build and test the evidence
Align measurements, compare interventions and validate predictions.
ROUTE 5 · CHAPTERS 17–20
Repair and transfer
Diagnose weak explanations, solve a rooftop case and continue learning.
Full chapter index · Try the transfer task · Gold Standard subject hub
Chapter index
Connect weather with exposure · Chapters 1–4
Follow exchange and timing · Chapters 5–8
Read the living response · Chapters 9–12
Build and test the evidence · Chapters 13–16
CHAPTER 1 OF 20 · Connect weather with exposure
1. Define the living system and the atmospheric question
The International Society of Biometeorology places plants, animals and humans within the field. The shared concern is interaction. Weather describes atmospheric conditions; biometeorology asks how those conditions meet a living system with particular structures, behaviour and history.
Begin by naming the response. Are we explaining leaf temperature, water loss, flowering date, animal activity or a measured human thermal environment? These are different outcomes. A project that records only air temperature cannot silently change its conclusion from “the courtyard was warmer” to “the plants grew less” without collecting biological evidence.
For the fictional garden, the first question is narrow: why are sunlit leaves warmer than shaded leaves at midday? The second asks whether the difference changes water loss. The third asks whether repeated differences affect development. Each step adds a new quantity and a longer time scale. Evidence for the first does not automatically answer the third.
Neighbouring subjects provide useful boundaries. Atmospheric Science explains the physical atmosphere. Ecology studies relationships among organisms and environments. Aerobiology follows biological material carried in air. Biometeorology connects atmospheric exposure with living responses and feedbacks; it does not replace those subjects.
Write a question that could be answered incorrectly. “Weather matters to plants” is too broad to test. “Under comparable soil-water conditions, the shaded plants have lower midday leaf temperatures than the exposed plants” identifies a comparison, a response and a condition that needs checking.
CHAPTER 2 OF 20 · Connect weather with exposure
2. Replace the weather-station number with an exposure pathway
A weather station, a leaf surface and a shaded bench occupy different physical positions. Height, nearby surfaces, obstacles and movement affect what each experiences. Microclimate means the local atmospheric environment at the scale relevant to the question; it does not mean a small weather station is automatically representative.
Imagine a regional station reports 30°C while two garden sensors read 29°C and 32°C. Before interpreting the difference, ask whether the garden instruments measure shaded, ventilated air or have themselves warmed in sunlight. Then ask whether their positions represent the leaves being studied. A correct reading at the wrong position can answer the wrong question very precisely.
Exposure is also a time history. A mobile animal may spend twenty minutes in a sunny patch, forty beneath vegetation and the rest elsewhere. A stationary sensor describes an available environment. It does not establish how long the animal occupied it. Location and behaviour are part of the link between environment and response.
For the garden, draw a simple inventory: regional air, air above the canopy, air beside a leaf, the leaf itself, soil and nearby paving. Specify which instrument measures each component. Keep leaf temperature distinct from air temperature throughout the record.
The specialist guide to atmospheric instrument siting and measurement develops this distinction. The practical repair for an apparent contradiction between sensors is to inspect exposure and measurement purpose before averaging the values into one reassuring number.
CHAPTER 3 OF 20 · Connect weather with exposure
3. Follow energy before predicting temperature
An object’s temperature changes when its energy storage changes. Radiation, exchange with moving air, contact with another surface and evaporation can all matter. A leaf can absorb sunlight while losing energy through evaporation and exchanging sensible heat with the air. “It is in the sun” identifies one input, not the complete balance.
For a simplified patch of vegetated ground, take net radiation as positive into the surface and sensible, latent and ground heat fluxes as positive away from it. Suppose net radiation is 420 watts per square metre, sensible heat is 110 and ground heat is 30. If storage and other terms are negligible, the residual latent heat flux is 420 − 110 − 30 = 280 W/m².
This is an invented balance using the framework described in FAO’s introduction to evapotranspiration. It is not a measured evaporation rate. An omitted storage term or a biased radiation measurement would be absorbed into the residual and mislabelled as latent heat.
Why a balanced equation can still mislead
Suppose actual storage is increasing at 40 W/m². The latent term becomes 240 W/m², not 280. The arithmetic was correct in both calculations; the system description changed. A budget is only as informative as its boundary, sign convention, time interval and included processes.
Also distinguish flux from temperature. A value of 280 W/m² describes energy transferred per area per second. It does not mean a surface is 280 degrees warmer, nor determine temperature without additional relationships. Begin with units whenever a physical explanation becomes confusing.
CHAPTER 4 OF 20 · Connect weather with exposure
4. Interpret humidity through vapour pressure
Relative humidity compares actual water-vapour pressure with saturation vapour pressure at the same temperature. Vapour pressure deficit, or VPD, is the difference between saturation and actual vapour pressure. FAO’s meteorological-data chapter explains these quantities and their role in evaporation calculations.
Consider two invented air samples. Use supplied saturation pressures rather than deriving them: 3.17 kPa at 25°C and 4.24 kPa at 30°C. Both samples have 60% relative humidity.
| Air temperature | Saturation pressure | Actual pressure at 60% RH | VPD |
|---|---|---|---|
| 25°C | 3.17 kPa | 1.902 kPa | 1.268 kPa |
| 30°C | 4.24 kPa | 2.544 kPa | 1.696 kPa |
The warmer sample has a larger deficit despite the identical percentage. This explains why a percentage alone can conceal a meaningful change in evaporative conditions. Yet VPD is not a direct measurement of plant water loss: available energy, water supply and resistance to transfer still matter.
The air-based deficit also differs from the gradient between a warm leaf’s internal evaporating surfaces and the surrounding air. If leaf temperature differs from air temperature, using air temperature for both sides changes the physical question.
For a field record, preserve temperature and humidity together, including their simultaneous timestamps. Calculating VPD from a daily maximum temperature and an unrelated humidity observation may construct a condition that never occurred. A tidy spreadsheet cannot repair mismatched inputs after their timing has been discarded.
CHAPTER 5 OF 20 · Follow exchange and timing
5. Connect boundary layers, stomata and water supply
Air movement near a surface differs from movement farther away. The thin region across which heat and vapour move helps determine transfer between a leaf and its surroundings. Leaf size, orientation and airflow therefore matter alongside the properties of the air itself.
A plant adds biological regulation. Stomata are openings involved in gas exchange, and their behaviour affects the pathway for water vapour. The FAO Penman–Monteith framework separates aerodynamic resistance from surface resistance. That separation prevents us from treating wind and plant behaviour as interchangeable controls.
Imagine a deliberately simplified transfer model: flux is proportional to a fixed driving difference divided by total resistance. If resistance falls from 200 to 100 arbitrary compatible units while the driving difference stays fixed, modelled flux doubles. But if the plant changes its stomatal behaviour at the same time, total resistance may not halve. If the leaf cools, the driving difference may also change.
The useful lesson is conditional reasoning. “More wind doubles water loss” is not supported by this example. “Halving the model’s total resistance doubles its predicted flux when the driving difference is held fixed” is supported. The conditions are part of the result.
In the garden, measure or constrain several pieces before choosing an explanation: leaf temperature, local airflow, soil-water condition and the plant’s developmental state. A dry-rooted plant and a well-watered plant may respond differently to the same afternoon atmosphere. The atmosphere creates an opportunity for exchange; the living system helps determine the realised exchange.
CHAPTER 6 OF 20 · Follow exchange and timing
6. Make water balances and energy balances agree
Water accounting asks where water enters, leaves or accumulates. Energy accounting asks what powers transfers such as evaporation. The two descriptions must be compatible, although neither alone identifies every pathway.
Take a fictional sealed-bottom experimental tray with no runoff or drainage. It receives 8 mm of water during a day and gains 3 mm in stored water. If evaporation and transpiration are the only remaining losses, their combined loss is 8 − 3 = 5 mm. If an unnoticed leak removes 2 mm, the atmospheric loss is instead 3 mm. The tray’s mass change cannot distinguish these pathways without the boundary assumption.
For another calculation, suppose 4.90 MJ/m² goes specifically into vaporising water. Using the approximate latent heat 2.45 MJ/kg at 20°C given by FAO, the evaporated mass is 2 kg/m², equivalent to about 2 mm of water. Energy used to warm soil or air is not included in that allocation.
From a model output to a practical question
Reference evapotranspiration describes a specified reference surface. It is not a direct reading of every plant’s water demand. Even a sound crop estimate must be connected to rainfall, stored soil water, root access and other losses before it becomes part of a management assessment.
Agronomy extends these distinctions to field crops. A useful learner checks three separate statements: what the atmosphere could support under a reference model, what the actual surface lost, and what changed in accessible water storage. Confusing those statements produces convincing numbers with incompatible meanings.
CHAPTER 7 OF 20 · Follow exchange and timing
7. Treat seasonal timing as a biological response
Phenology concerns recurring events such as leaf emergence, flowering and insect development. A calendar date describes when an event was recorded. An explanation needs to connect that event with biological state and environmental history.
Temperature accumulation can be useful, but it is not a universal clock. Some seasonal responses involve prior chilling, daylength and other cues. In a 2020 controlled study of six temperate woody species, Wang and colleagues varied chilling, forcing temperature and photoperiod. The influences and interactions differed among species. The result supports testing multiple cues rather than assigning every plant the same response to warmth.
Our garden question must therefore name an event and organism. “Spring arrived” is too vague. “The first open flower on each of twelve marked individuals, under a stated observation protocol” is measurable. A model for first flowering need not describe fruit development or the timing of another species.
History also changes the meaning of a temperature reading. Two plants at the same temperature today may have experienced different preceding weeks. If the response depends on accumulated conditions or a completed developmental stage, today’s temperature cannot stand in for the whole history.
Avoid transporting a temperate spring model into a tropical planting simply because both contain plants. Establish the relevant cues, species, starting event and local validation first. A model can be mathematically convenient and biologically inappropriate. The next chapter shows how to use a simple thermal-time calculation while preserving that distinction.
CHAPTER 8 OF 20 · Follow exchange and timing
8. Calculate growing degree days without making a universal promise
A simple growing-degree-day calculation subtracts a chosen base temperature from a daily mean and sets negative contributions to zero. The USA National Phenology Network explains accumulated growing degree days and their use in predicting some seasonal transitions.
For this teaching exercise, define daily mean as the average of the minimum and maximum, use a fictional base of 10°C, and assume the model’s starting event has already occurred. No upper-temperature adjustment is used.
| Day | Minimum | Maximum | Mean | Daily degree days | Running total |
|---|---|---|---|---|---|
| 1 | 6°C | 12°C | 9°C | 0°C·days | 0 |
| 2 | 8°C | 16°C | 12°C | 2°C·days | 2 |
| 3 | 10°C | 20°C | 15°C | 5°C·days | 7 |
| 4 | 12°C | 22°C | 17°C | 7°C·days | 14 |
| 5 | 10°C | 18°C | 14°C | 4°C·days | 18 |
A hypothetical threshold of 15°C·days is first exceeded in the daily record on day 5. The table does not locate the event’s exact hour. Nor does a predicted threshold crossing prove an observed flower opened.
Changing the base changes the model. With an 8°C base, the same means contribute 1, 4, 7, 9 and 6, totalling 27°C·days. One cannot keep the original threshold and claim to have improved the forecast merely by choosing the larger total.
The University of California’s archived explanation of phenology models emphasises species-specific developmental thresholds and checking models against local observations. Preserve the base, starting rule, temperature source, calculation method and any upper cutoff together. A total without those details is an incomplete measurement description.
CHAPTER 9 OF 20 · Read the living response
9. Separate the event from the first time somebody saw it
An organism can change between visits. Suppose a marked plant has no open flowers on Monday and has one on Thursday. If the observations are reliable, onset lies after Monday’s visit and no later than Thursday’s. Thursday is the first detection date, not necessarily the biological onset date.
Now compare a second plant visited every morning. Its recorded first flower may be earlier simply because the observation interval is shorter. A change in volunteer effort could create an apparent long-term shift even if true timing remained unchanged.
The existing guide to phenological first-event dates and detection lag examines this problem. Use repeated observations of marked individuals, a consistent event definition and explicit records of absences as well as presences. An empty entry should not silently become “event absent”.
Time of day can also alter what is observable. Observation effort, time-of-day windows and phenology mismatch provides the deeper route. A count of active insects at noon and a count at dawn differ in observation opportunity as well as weather.
For the garden report, retain interval bounds, visit times and missed visits. Where exact onset is unknowable, state the interval. Honest resolution is more useful than a date with false precision. The same reasoning applies when comparing manual observations with cameras: additional images improve temporal coverage, but only if classification and viewing conditions remain dependable.
CHAPTER 10 OF 20 · Read the living response
10. Include animal behaviour in the exposure model
Animals can change their exposure by moving, altering posture or changing activity. Different species also produce and exchange heat differently. A useful explanation therefore connects the available environment with the environment selected and the response observed.
Consider a fictional animal observed for one hour: fifteen minutes in an exposed patch at 34°C air temperature, thirty minutes in cover at 28°C and fifteen minutes elsewhere at 30°C. The time-weighted air temperature is (15 × 34 + 30 × 28 + 15 × 30) / 60 = 30°C.
That number summarises occupied air temperatures under the stated sampling. It is not body temperature, and it suppresses the sequence and extremes. The animal could have entered the exposed patch briefly several times or remained there continuously. Those histories share a mean while potentially differing in energy storage and opportunity for recovery.
The calculation also omits radiation, moisture, airflow and contact surfaces. It illustrates why an exposure summary must be named precisely. A map of shaded habitat does not demonstrate use; a movement trace does not by itself measure physiological consequence.
To study the connection, synchronise location observations with environmental measurements and define an outcome such as activity duration or body-temperature measurements collected under an appropriate research protocol. Record what could not be observed. Avoid inferring absence from a quiet period when the animal may be concealed.
This is a central biometeorological move: behaviour belongs inside the explanation. Treating the organism as a fixed sensor can remove the very response that allows it to navigate a variable environment.
CHAPTER 11 OF 20 · Read the living response
11. Distinguish human thermal conditions from a person’s response
For people, air temperature is only one part of thermal exchange. Radiation, humidity, airflow, activity and clothing affect the relationship between environment and body. Environmental measurements and indices describe selected aspects of that relationship; they do not measure every person’s physiological state.
The US National Weather Service’s heat-tools guidance explains that Heat Index combines air temperature and relative humidity and was devised for shade with light wind. Its discussion of other tools shows why the intended setting matters. An index should be read with its assumptions and purpose, not treated as an interchangeable label for every outdoor location.
In the fictional courtyard, two benches have the same measured air temperature. One receives direct sunlight and radiation from nearby warm surfaces; the other is shaded. The air-temperature equality does not establish equal radiant conditions. A project evaluating the benches should measure the variables needed for its chosen thermal assessment, not infer success from a single shaded thermometer.
Keep the conclusion matched to the study
Suppose a screen reduces a measured radiant indicator while air temperature changes little. The supported conclusion concerns the measured environment. Establishing comfort requires an appropriate assessment of people, and establishing health effects requires a different level of evidence. Neither follows automatically from the screen’s installation.
For a practical environmental report, state the index, input instruments, exposure positions, time interval and population assumptions. Use applicable local heat guidance for operational decisions. The educational task here is to understand what an index includes and leaves out, rather than invent a universal safe temperature from the garden calculations.
CHAPTER 12 OF 20 · Read the living response
12. Follow the feedback from organisms to atmosphere
The relationship runs in both directions. Vegetation changes the surfaces that absorb and reflect radiation, exchanges water with air, creates shade and modifies airflow. These changes can alter conditions for other organisms beneath and beside it.
In a 2019 study comparing forest and open-site temperatures, De Frenne and colleagues used paired measurements at 98 sites across five continents. Their results showed buffering beneath forest canopies, with cooler conditions when surrounding temperatures were hot and warmer conditions when they were cold. This is evidence about canopy environments, not a fixed cooling allowance for any newly planted tree.
Our garden can test a narrower claim. Measure matched positions under and outside a canopy through the day and across different weather. Examine maximum, minimum and duration statistics separately. A canopy that reduces afternoon maxima need not lower every hourly value by the same amount.
An edge adds another complication. Conditions can vary continuously from exposed paving through the canopy boundary into deeper cover. The guide to ecological edge effects and boundary leakage explains why a mapped border should not be treated as a physical wall.
A good feedback explanation names both the change and its reach: which surface changed, which atmospheric variable responded, how far the effect extended and under which conditions it appeared. “Greenery cools everything” loses the mechanisms needed to design or evaluate a real place.
CHAPTER 13 OF 20 · Build and test the evidence
13. Build a measurement record that can survive inspection
A biometeorological dataset needs two aligned records: environmental conditions and biological observations. If their clocks, locations or definitions differ, a statistical relationship may be an artefact of joining incompatible records.
For every environmental series, preserve sensor identity, units, measurement interval, location, height, shielding, calibration history and maintenance events. For each organism, preserve identity, position, developmental stage, observation method and relevant changes in treatment or surroundings. These details explain what a row of numbers actually represents.
Suppose one logger timestamps the start of each fifteen-minute interval while another timestamps the end. Joining identical timestamps shifts their physical windows. During a rapid change in cloud cover, this can make a leaf appear to respond before the radiation changed. Correct the time convention before interpreting a biological lag.
Missing data need reasons where available. A sensor that fails during heavy rain has a different missingness pattern from random communication losses. Calculating a monthly mean from the remaining dry intervals can distort the environment represented by the record.
A useful field-sheet structure
- Identify the organism, plot and sensor separately, then record how they are linked.
- Use a consistent time reference and preserve the original timestamp convention.
- Record visits, absences, uncertain observations and missing entries distinctly.
- Flag repairs, relocation, pruning, irrigation and changes to nearby surfaces.
- Retain raw measurements alongside documented corrections and derived quantities.
The purpose is reproducibility at the level of meaning. Another reader should be able to reconstruct why an observation was included, how it became an exposure estimate and which biological response it was compared with.
CHAPTER 14 OF 20 · Build and test the evidence
14. Test a shade intervention with a comparison that earns its claim
The garden team installs shade above one plot and sees its afternoon leaf temperature fall. Weather also changes that week. A before-and-after difference combines the intervention with everything else that changed.
Consider this invented comparison of mean leaf temperatures during matched afternoon windows:
| Plot | Before | After | Change |
|---|---|---|---|
| Shaded intervention plot | 34°C | 31°C | −3°C |
| Comparison plot | 33°C | 32°C | −1°C |
The difference between changes is −3 − (−1) = −2°C. Under the assumption that the intervention plot would otherwise have followed the comparison plot’s change, this estimates a two-degree reduction associated with the intervention beyond the shared change.
That assumption is substantive. If the intervention plot alone was irrigated, if its sensor was moved or if the plots respond differently to the week’s weather, the estimate may combine several effects. The arithmetic cannot establish the counterfactual assumption.
Repeated measurements improve the description of each plot, but a thousand timestamps from one treated plot do not become a thousand independently treated plots. Replicate the intervention across suitable units when the question requires a broader conclusion. Randomise allocation where feasible and keep comparison conditions documented.
Also specify what the intervention is. A shade screen may alter airflow and rainfall interception as well as radiation. If the intended claim concerns the whole screen, those changes belong to its effect. If the claim concerns radiation alone, they become alternative pathways to measure or control. Good experimental design begins with the claim, not with the statistical test selected afterward.
CHAPTER 15 OF 20 · Build and test the evidence
15. Evaluate predictions on genuinely new conditions
A model fitted to the garden’s observations may reproduce them well because it has learned the site’s particular pattern. That does not establish reliable prediction at a different garden, in a later season or after a change in management.
Separate fitting from evaluation. If the intended use is next season, preserve later observations for testing. If the intended use is another site, evaluate on sites not used for fitting. Randomly mixing adjacent timestamps between training and test sets can give the model near-duplicates of the conditions it is supposed to predict independently.
For a fictional flowering model, predictions for three held-out plants are days 20, 24 and 28. Observed event dates are 22, 23 and 31. With error defined as prediction minus observation, errors are −2, +1 and −3 days. Mean error is −4/3 days, about −1.33, and mean absolute error is (2 + 1 + 3)/3 = 2 days.
The signed mean reveals a tendency toward early predictions in this tiny example; absolute error describes typical magnitude without cancellation. Three cases provide little evidence about performance across species or climates. If observations themselves are intervals, exact-date scoring needs additional care.
Compare the model with a simple baseline, such as a historically typical date appropriate to the same event and population. A complicated model is useful only if its additional information improves the task under a fair test.
Finally, report where the model has not been evaluated. Missing validation in unusually hot, dry or shaded conditions is a limit on use, even if those conditions fit neatly into the software’s input form.
CHAPTER 16 OF 20 · Build and test the evidence
16. Keep weather responses, climate change and adaptation distinct
A warm afternoon, a warm season and a changing long-term climate distribution operate on different time scales. A short experiment can reveal a mechanism without estimating every consequence of a long-term shift.
Suppose the garden’s flowers appear earlier in a warmer year. That association may be consistent with thermal-time reasoning. To attribute a persistent trend, examine the observation method, species composition, management, site changes and a sufficiently informative environmental record. A single pair of years cannot separate all these possibilities.
Future conditions may also change combinations rather than one variable alone. Warmer air can coincide with altered water availability, different cloud conditions or changed extremes. A model fitted where these variables moved together may struggle when their relationships change.
Distinguish an immediate response from acclimatisation and evolutionary adaptation. A short-term change in activity is not evidence that a population has genetically adapted. Demonstrating adaptation requires evidence about heritable variation and selection across generations, beyond the temperature measurements in this example.
Scenario reasoning is useful when its assumptions remain visible. Ask what the current model predicts if a specified input changes, then identify which parameters or biological relationships might also change. Present that result as conditional. The valuable output is a map of dependencies and uncertainties, not a single exact date presented as the inevitable future of every organism at the site.
CHAPTER 17 OF 20 · Repair and transfer
17. Diagnose a failed explanation before collecting more of the same data
When measurements disagree with a prediction, additional data help only if they address the weak link. Locate the mismatch in the chain from atmosphere to exposure, mechanism, response and observation.
| Symptom | Possible weak link | Useful next check |
|---|---|---|
| Leaf temperature exceeds reported air temperature | Different physical quantities or radiation exposure | Compare leaf and properly measured local air temperatures |
| High VPD accompanies low measured water loss | Water supply, resistance or measurement boundary | Check soil-water status and whether the loss estimate includes all pathways |
| Flowering appears suddenly earlier | Observation effort or changed individuals | Audit visit intervals, event definitions and plant identities |
| A model succeeds at one site and fails elsewhere | Changed exposure or model relationship | Compare input ranges, siting and biological conditions |
| Shade appears effective only after averaging | Timing or cancellation of opposite effects | Inspect matched hourly windows and distributions |
Do not treat every unusual measurement as instrument error. A real local extreme may be exactly what the organism experienced. Conversely, do not protect an attractive biological story by ignoring a damaged radiation shield or an incorrectly joined timestamp.
A practical repair record has four sentences: what failed, which assumption might explain it, what observation would distinguish alternatives, and how the conclusion will change under each result. This keeps troubleshooting connected to the scientific question.
For the garden, the next useful action might be one day of paired, well-described measurements rather than another month from the same poorly placed sensor. Data volume cannot substitute for observing the missing relationship.
CHAPTER 18 OF 20 · Repair and transfer
18. Transfer the reasoning to a rooftop planting
The garden team now investigates a rooftop. It proposes reusing the garden’s thermal-time model and shade conclusion. Work through the new evidence before deciding what transfers.
The rooftop’s five daily mean air temperatures are 11, 14, 16, 9 and 15°C. Use the same fictional base of 10°C and zero negative daily contributions. A hypothetical model threshold is 14°C·days from the stated starting event.
At one matched instant, the rooftop air has a supplied saturation vapour pressure of 4.00 kPa and relative humidity of 50%. The garden air has saturation pressure 3.20 kPa and relative humidity of 75%. Calculate each air-based VPD.
Finally, before and after a shade installation, the rooftop intervention plot’s mean leaf temperature changes from 35 to 32°C. Its comparison plot changes from 34 to 30°C. Calculate the difference between changes. Does the simple before-and-after result show an additional cooling effect from shade?
State three reasons the original garden model might need new validation, and distinguish the model’s predicted threshold crossing from a measured biological event.
Answer and reasoning
Daily thermal contributions are 1, 4, 6, 0 and 5°C·days. The running total is 1, 5, 11, 11 and 16. The hypothetical threshold is first exceeded in the daily record on day 5. This does not identify an exact event time or prove the organism reached the predicted stage.
Rooftop actual vapour pressure is 4.00 × 0.50 = 2.00 kPa, giving VPD 2.00 kPa. Garden actual pressure is 3.20 × 0.75 = 2.40 kPa, giving VPD 0.80 kPa. The rooftop deficit is 2.5 times as large. That ratio is not automatically the ratio of plant water loss.
The intervention change is −3°C and the comparison change is −4°C. Their difference is +1°C. Under the comparison assumption, the intervention plot cooled one degree less than expected from the comparison plot. The raw three-degree fall therefore does not establish additional cooling due to shade. Investigate uncertainty and other differences before concluding that the screen caused warming.
New radiation exposure, wind, rooting conditions, plant history or species identity could each undermine transfer. The transferable achievement is the method of asking and testing, rather than the unexamined reuse of a fitted number.
CHAPTER 19 OF 20 · Repair and transfer
19. Answer the questions that most often cause confusion
Is biometeorology just another name for meteorology?
The fields overlap, but biometeorology places living responses and feedbacks at the centre of the question. A forecast of local temperature becomes part of a biometeorological investigation when it is connected to an organism’s exposure, physiology, behaviour or development.
Does lower air temperature always mean less thermal stress?
Air temperature alone is insufficient. Radiation, moisture, airflow and the organism’s condition affect exchange. Even the word “stress” needs a defined biological response. A cooler air measurement can be useful evidence without establishing the complete outcome.
Are growing degree days measurements or predictions?
They are derived quantities calculated from temperature data under a specified method. Using them to estimate an event requires a biological model and a threshold appropriate to that model. Observing the event is a separate step that allows the prediction to be checked.
Can a satellite surface temperature replace a local air sensor?
They measure different things. A surface-temperature estimate may be valuable for the relevant surface and viewing conditions, but it is not automatically air temperature beside a leaf or body temperature. Establish the measurement relationship before substitution.
Is a shaded sensor biased?
It depends on the intended quantity and installation. Shielding an air-temperature instrument from direct radiation can be essential. Placing a sensor beneath a canopy also selects a local environment. Instrument protection and representativeness are separate questions.
Why did the plants differ when they received the same water?
Equal water delivered does not prove equal water available to roots. Storage, drainage, rooting, evaporation and plant size may differ. Start with the water balance and the actual root environment rather than assuming the watering record fully describes plant water status.
Does a successful prediction prove the proposed mechanism?
No. Different explanations can predict similar observations within a limited dataset. Tests that vary relevant conditions, collect missing quantities or distinguish competing pathways provide stronger evidence about mechanism than repeated agreement under one familiar condition.
CHAPTER 20 OF 20 · Repair and transfer
20. Build a practical learning route and an evidence trail
Begin with an observable comparison that does not require exposing organisms to harmful extremes: shaded and exposed environmental positions, repeated observations of naturally occurring flowering, or a supplied dataset. Define the question, units and observation schedule before collecting values.
For a first exercise, calculate a five-day thermal sum and explain why the base and start date matter. Next, compare simultaneous temperature and humidity records using VPD. Then interpret a paired shade comparison, including the assumptions needed to attribute an effect. These steps connect arithmetic with biological meaning.
A stronger project produces a short evidence record: a map of positions, an instrument and organism register, raw observations, a documented calculation, a graph or table of the comparison, and a conclusion that names its limits. Keep the original data so another learner can challenge the interpretation without reconstructing the entire field visit.
Continue through the connected subjects
Use Agronomy for crop, soil and water decisions; Ecology for populations and living relationships; and Atmospheric Science for radiation, air and moisture. Return to How X Works to follow the broader method of connecting parts, processes, evidence and failure points.
Technical reading behind the guide
- International Society of Biometeorology: the field and its questions establishes the subject boundary.
- FAO Irrigation and Drainage Paper 56, especially chapters 1–3 linked above, provides the physical framework and measurement definitions.
- USA National Phenology Network: accumulated growing degree days explains the thermal-time product and calculation.
- Wang and colleagues, 2020 provides an experimental example of interacting phenological cues.
- De Frenne and colleagues, 2019 provides evidence about canopy microclimate buffering.
- National Weather Service heat tools explains the purposes and assumptions of environmental heat indices.
The subject becomes useful when a reader can trace a claim all the way from atmospheric conditions to the organism and back to the evidence. That trace makes it possible to improve a measurement, repair an explanation and recognise which conclusions remain dependable when the setting changes.