The short answer
Air VPD (VPDair) takes saturation vapor pressure at air temperature; leaf VPD (VPDleaf) takes it at leaf temperature. Both subtract the same room vapor pressure. Most room readouts show VPDair, a valid measure of how dry the air is. Transpiration follows a different gap, between the nearly saturated air inside the leaf and the room, so VPDleaf is the deficit the plant works against. The two match only when leaf and air temperature match. A warmer leaf sees a bigger deficit than the dashboard shows, a cooler leaf a smaller one. CO₂ enrichment, water stress and still air push leaves warmer.
Key takeaways
- VPDair and VPDleaf share the same ambient vapor pressure; only the saturation side changes, taken at air or at leaf temperature.
- In Kubota's 2023 worked example at 20 °C air and 60% RH, VPDair is 0.94 kPa, while a leaf at 22 °C works against a VPDleaf of 1.24 kPa.
- Nelson and Bugbee's 2015 energy-balance model put unstressed leaves typically within 2 °C of air, with modeled gaps of 6 to 12 °C only under water stress and low wind.
- In a single leaf-chamber study of four cannabis varieties, Chandra et al. (2011) found 700 versus 390 μmol mol−1 CO₂ cut stomatal conductance 23–44% depending on variety.
- Label every VPD figure as air or leaf, and read upper and lower canopy leaves before moving a setpoint.
The short video version of this article
What VPDair and VPDleaf each calculate
Kubota (2023), writing in the e-GRO Edible Alert extension series, puts the two calculations side by side:
- VPDair = SVPair − VPair
- VPDleaf = SVPleaf − VPair
VPair is the vapor actually in the room, worked out from air temperature and RH (VPair = SVPair × RH/100). Only the saturation term changes: taken at the temperature your sensor reads, or at the temperature of the leaf.
Most room VPD readouts use air temperature for saturation. That is a legitimate number: it describes the drying power of the air, and air temperature and RH are what your HVAC answers to. It is not the deficit across the stomata, which is why Kubota recommends labelling which one you mean. The stage targets in a standard cannabis VPD chart are VPDair numbers, so they describe the room, not the leaf.
Why transpiration follows leaf temperature
Kubota (2023) describes transpiration from a single leaf as proportional to the vapor pressure difference between the stomatal cavities and the surrounding air. Those cavities are treated as nearly 100 % RH, and saturation vapor pressure is set by temperature, so the vapor inside the leaf depends on leaf temperature, not on the wall sensor.
Cernusak et al. (2024), in a Plant, Cell & Environment review, trace the assumption to Gaastra (1959): intercellular vapor pressure is taken as saturation at leaf temperature, and every commercial portable photosynthesis system uses it to calculate stomatal conductance. Nelson and Bugbee (2015) drive latent heat loss in their leaf energy-balance model with the same deficit, paired with vapor conductance.
That pairing matters. Water loss depends on the deficit and on how open the stomata are, so a higher VPDleaf only means more transpiration if conductance holds.
How far leaves run from air temperature
Kubota's (2023) worked example shows the scale. At 20 °C air and 60% RH, VPDair is 0.94 kPa; put the leaf at 22 °C, 2 °C above the air, and VPDleaf is 1.24 kPa. Same thermostat and RH, drier conditions at the leaf. With a leaf cooler than the air, VPDleaf comes out below VPDair and the setpoint is wetter at that leaf.
Nelson and Bugbee (2015) modeled leaf temperature under field sun, LED, greenhouse and HPS radiation. With plants not water stressed, leaves in all four sources typically stayed within 2 °C of air. The large offsets needed water stress and low wind together, and then the radiation source mattered: the model's greenhouse case ran hotter than LED, so a greenhouse or light-dep house has a higher ceiling than an LED room when transpiration stalls.
| Scenario | Leaf vs air | Model conditions |
|---|---|---|
| All four radiation sources | Typically within 2 °C | Not water stressed |
| Every radiation scenario | +1 °C | Conductance cut 30% by elevated CO₂ |
| Field | +6 °C | Water stress, low wind |
| LED | +8 °C | Water stress, low wind |
| Greenhouse | +10 °C | Water stress, low wind |
| HPS | +12 °C | Water stress, low wind |
What CO₂ enrichment does to leaf temperature
Nelson and Bugbee (2015) note that supplemental CO₂ in controlled environments can reduce stomatal conductance about 10–40%, citing earlier work. Less conductance means less evaporative cooling. In their model, a 30% drop raised leaf temperature 1 °C in every radiation scenario tested: one modeled case, not a conversion factor for your room.
The cannabis data come from Chandra et al. (2011), a single leaf-chamber study on upper leaves of four Cannabis sativa varieties in a cuvette at 25 °C, with CO₂ raised from 390 to 700 μmol mol−1. In that study conductance fell 44%, 43%, 30% and 23% in HPM, K2, MX and W1, and transpiration fell 45%, 44%, 50% and 51%. It had no canopy VPD, flower temperature or yield data.
Two things follow. The same CO₂ step closed stomata by very different amounts across varieties, so one leaf check per room may miss some cultivars. And because water loss pairs deficit with conductance, a CO₂-warmed leaf can show a higher VPDleaf while moving less water.
Drop in stomatal conductance at 700 vs 390 μmol mol−1 CO₂, by cannabis variety
- HPM44%
- K243%
- MX30%
- W123%
Do cannabis flowers run hotter than leaves?
Nelson and Bugbee (2015) did not measure flowers. They argue their near-worst-case leaf scenario would likely represent thick, dense, low-transpiring tissue, naming Cannabis flowers alongside tomatoes and strawberries.
The reasoning has two parts. Dense tissue transpires little, so it sheds little heat by evaporation. It also absorbs more shortwave: in their analysis only 63% of HPS shortwave radiation is absorbed by the first leaf, while a structure ten times thicker would absorb more than 80%.
In late flower, that points to dense tops under strong light and still air as where VPDleaf most likely departs from the room reading. That is a hypothesis from a leaf model, not a measurement. Water stress was one of the two conditions behind the large modeled gaps, which makes irrigation timing part of leaf-temperature control. Still air was the other: airflow through the canopy, not room air exchange.
The limit: substomatal air is not always saturated
VPDleaf assumes the stomatal cavities are saturated at leaf temperature. That holds closely, not perfectly. Wong et al. (2022), measuring gas exchange on leaves generally rather than cannabis, found substomatal relative humidity near 97% under a mild leaf-to-air humidity difference and around 80% when the difference was large. Cernusak et al. (2024) review how that unsaturation biases the conductance estimates gas-exchange instruments report.
So VPDleaf overstates the real gradient most when the leaf-to-air difference is largest, which is when you would lean on it hardest. It still moves when the leaf warms or cools, which VPDair cannot do. Use it for direction and for comparing zones or treatments, not for precision beyond what the saturation assumption allows.
How to measure and steer with VPDleaf
- Label every number. Mark dashboards, SOPs and target sheets as VPDair or VPDleaf, as Kubota (2023) recommends. A target copied across is off by the leaf offset.
- Read upper and lower leaves separately. Kubota notes leaf temperature differs across the canopy, upper versus lower leaves, while all of them contribute to transpiration.
- Compute it one way. SVP at leaf temperature, minus VPair from air temperature and RH, with the air reading taken in the same zone as the leaf.
- Re-measure when CO₂ goes on, cultivar by cultivar. Nelson and Bugbee's model and Chandra's cuvette data both point to enrichment shifting leaf temperature, not the air reading, by amounts that varied by variety.
The controller can stay on VPDair. Use VPDleaf as the check on what that setpoint means at the leaf, and change the setpoint only when your own leaf readings show the offset has moved.
Frequently asked questions
Is VPDleaf always higher than VPDair?
No. A leaf cooler than the air gives a VPDleaf below VPDair, which Kubota (2023) also shows. A leaf temperature logged without the air temperature beside it cannot tell you which way the gap runs.
Should I lower my VPD setpoint when I run CO₂?
None of these studies gives a new setpoint. Log leaf temperature in the same room with enrichment off and on, recompute VPDleaf for both, and move targets from that comparison.
Why does my leaf temperature read almost the same as the air?
With well-watered plants and moving air, that matches Nelson and Bugbee's (2015) finding that unstressed leaves typically sat within 2 °C of air. Even that range moved Kubota's example from 0.94 to 1.24 kPa, so watch for drift after CO₂ goes on or an irrigation is missed.
Does LED or HPS change how hot leaves get?
In Nelson and Bugbee's (2015) model it mattered mainly under stress, where worst-case leaves rose 8 °C over air under LED and 12 °C under HPS. The light source sets how bad a stall can get more than the everyday offset.
Do the Chandra results show which cultivars handle CO₂ best?
No. The ranking only describes how far upper-leaf stomata closed at one CO₂ step in a cuvette, with no yield or flower data. It is a reason to check leaf temperature per cultivar, not a cultivar recommendation.
Sources
- VPDleaf vs. VPDair - Two different ways to determine VPD · Kubota, 2023, e-GRO Edible Alert (Ohio State University-led extension consortium)
- Unsaturation in the air spaces of leaves and its implications · Cernusak et al., 2024, Plant, Cell & Environment
- Analysis of Environmental Effects on Leaf Temperature under Sunlight, High Pressure Sodium and Light Emitting Diodes · Nelson & Bugbee, 2015, PLOS ONE
- Photosynthetic response of Cannabis sativa L., an important medicinal plant, to elevated levels of CO2 · Chandra et al., 2011, Physiology and Molecular Biology of Plants
- Humidity gradients in the air spaces of leaves · Wong et al., 2022, Nature Plants
Know what is on the bench before you compare leaves
If your leaf readings differ by cultivar, the comparison only holds when every plant is what its tag says. Prime Cuts Nursery is a licensed Sonoma County nursery supplying rooted clones and teens; menu strains carry documented breeder lineage, and every mother is RT-qPCR tested for HLVd and Fusarium on a recurring schedule.
How we test mothers for HLVd · Call or text (707) 494-5197 · Mon–Fri 9am–5pm PT · pickup at the nursery by appointment.
