Does VPD need leaf temperature, or is air close enough?

The short answer

VPD is a statement about how hard the air is pulling water out of a leaf. That calculation needs the temperature of the surface the water is leaving from. Air temperature is a stand-in for it, and the substitution is worse than the VPD charts suggest. Not because it’s off by some steady amount you could just subtract — because how far off it is keeps changing, and so does which way it’s off.

What the difference looked like on my own tent

One morning last August my tent was steered on air-derived VPD and read 1.035 kPa — comfortably inside the veg target band. Eight minutes later the same tent, measured off the leaf surface with an infrared sensor, read 0.612 kPa: below band.

That is a gap of 0.42 kPa, which is roughly the width of the entire veg band. One number said “you’re fine, hold here.” The other said “you’re under, do something.” Same tent, same morning, same plants.

That is one tent on one morning and I’m not going to dress it up as a general constant — that would be exactly the mistake this page is about. What it is: an existence proof that the two numbers can disagree by enough to change the decision.

Why you can’t just subtract a couple of degrees

That’s the tempting patch — knock two degrees off for LED, carry on. I believed a version of it myself and had it written into my own notes, right up until I went and read the paper it was supposedly derived from.

The best measurements I know of are Nelson & Bugbee (2015, PLOS ONE), who compared leaf and air temperature under sunlight, greenhouse light, HPS and LED. Their findings, in short:

• For plants that are not water stressed, leaves sat within about 2 °C of air temperature under all four light sources.
• Swapping HPS for LED moved leaf temperature by only around 1.3 °C.
• The effect of the plant’s water status and its own evaporative cooling was much larger than the effect of the light source.
• Under water stress with high light and low airflow, leaves ran hotter than air — by around 8 °C under LED, and more under HPS.

Read those together and the lamp-offset idea falls apart. The gap has very little to do with your light and almost everything to do with the plant. A leaf keeps itself cool the same way you do — by sweating. Water goes out through the pores on the underside, evaporates, and carries the heat off with it. A plant that’s drinking freely sweats freely and sits right about at air temperature. A plant that’s short of water closes those pores to hang on to what it has, stops sweating, and then just sits there under the lamp getting warmer.

So the gap is small while everything is fine, and it gets big — and turns around — right when it isn’t. The leaf goes from a touch cooler than the air to several degrees hotter than it. That’s the part that kills the correction factor: knock two degrees off for your LED, and on the day the plant is actually in trouble you’re now ten °C out, in the wrong direction, on the one reading you needed to be right. Air VPD is least trustworthy at exactly the moment it matters most.

What this means if you don’t own an IR sensor

Buy a handheld infrared thermometer. They’re about twenty dollars. Point it at a healthy fan leaf in the canopy, compare to your air temperature, and recompute VPD off the leaf number.

This isn’t a fringe opinion. AC Infinity — who make the controllers a great many hobby tents run on — sell a handheld IR thermometer for exactly this, so growers can check leaf VPD and hand-calibrate their controller against it. The largest player in the category agrees leaf temperature is the right number.

The honest caveat on the handheld approach is the same physics as above: a spot check tells you the gap at that moment, and the gap moves with the plant’s water status through the day. It’s a real improvement over ignoring the leaf. It is not the same as watching it continuously.

Where my device fits, stated narrowly

Croplock reads leaf-surface temperature from an infrared sensor in the canopy, and that leaf reading is what its control loop runs on — not a VPD worked out from the air. Plenty of gear can adjust your tent on its own now; the difference is which number it is adjusting toward. When it can’t get a trustworthy leaf reading — sensor unavailable, canopy too sparse to be sure it’s looking at a leaf — it falls back to plain air VPD and records that it did so, rather than inventing a leaf it can’t see. I removed the fixed-offset fallback it used to have, for the reason above: one number can’t correct for a gap that runs cool some of the time and hot the rest.

Sources

• Nelson, J.A. & Bugbee, B. (2015). Analysis of environmental effects on leaf temperature under sunlight, high pressure sodium and light emitting diodes. PLOS ONE 10(10): e0138930.
• Bench measurement, 2026-08-02, from my own tent’s logs.

More of these at Answers. The device is at croplock.com.