Many growers ship great harvests without ever looking at VPD. That’s fine. VPD is a tool, not a religion. If you do use it, you’ll get better results by basing it on leaf temperature rather than air temperature, because leaves are almost always cooler than the surrounding air when they’re transpiring.

What VPD actually measures (and why leaf temperature matters)

Vapor Pressure Deficit (VPD) is how “thirsty” the air is, expressed in kPa. It’s the gap between:

• the SVP at the leaf surface temperature: how much water vapor the boundary layer could hold if fully saturated, and
• the AVP of the room air, which depends on the air’s temperature and relative humidity.

Leaves drive transpiration, not the room air. Under light, water evaporating from the leaf cools it (evaporative cooling), so leaf temperature (T_leaf) is often a couple degrees cooler than air temperature (T_air). Using T_air in a VPD calculator overestimates VPD and can push you to run drier or hotter than the plant wants.

The math (kept practical)

Let temperature be in °C and RH in %.

Saturation vapor pressure (SVP) in kPa (Tetens approximation):

SVP(T) = 0.61078 × exp((17.27 × T) ÷ (T + 237.3))

Actual vapor pressure (AVP) comes from the air:

AVP = (RH ÷ 100) × SVP(Tair)

Leaf-based VPD:

VPDleaf = SVP(Tleaf) – AVP

…not just (1 – RH) × SVP(Tair).

Why this changes your targets (worked examples)

A 2 °C cooler leaf chops ~0.35–0.40 kPa off the “air-only” VPD—big enough to be the difference between happy and “tacoing”.

Rule of thumb: If you don’t measure leaf temperature, assume leaves run ~1–3 °C cooler than air under strong light, or better, measure and use the real offset.

How to measure leaf temperature well

You don’t need a thermal camera. A basic IR “laser” thermometer works if you respect its limits.

1. Measure multiple leaves at canopy height (center, corners, mid-edges). Record 6–10 readings and average.
2. Shade the spot briefly with your hand or clipboard to avoid reading glare from LEDs.
3. Angle matters: read at ~45° to the leaf to reduce specular reflections.
4. Emissivity: most IR guns assume 0.95; leaves are ~0.95–0.98, so you’re close enough.
5. Time of day: re-check after big events—lights on, mid-cycle, post-watering, pre-lights-off—because transpiration changes leaf temperature.

Now measure air temperature (at canopy level) with the same device family if you can, or at least the same place and height.

Compute the leaf temperature offset (LTO)

Define LTO = Tleaf – Tair

Example: air = 28 °C, leaf = 26 °C → LTO = -2 °C.

• Many controllers let you input Leaf Temperature Offset so their VPD readout compensates. If your leaf is cooler, you enter a negative number.
• If your controller doesn’t support it, compute VPD yourself with the equations above (use T_leaf for SVP on the left term, and T_air for AVP on the right).

Consistency beats precision: use the same method every time, log the offset, and track how it drifts after watering, defoliation, and through the stretch.

Targets that behave in the real world

These are leaf-based VPD ranges that keep stomata open without inviting mildew (assumes reasonable light and nutrition):

• Seedling: 0.6–0.8 kPa
• Veg: 0.8–1.2 kPa
• Early flower: 1.1–1.3 kPa
• Mid/late flower: 1.3–1.6 kPa

Translate those into temperature/RH setpoints with your preferred chart or app, but verify with leaf-based VPD once the room stabilizes (30–60 minutes).

An easy to read reference chart

Practical control tactics

• Humidifier/dehumidifier sized for the room. Check liters/day, not watts.
• Smooth air exchange (low, constant) instead of hard on/off cycling.
• Airflow across the canopy, not into leaves—avoid windburn, thin the boundary layer, and help CO₂ distribute.
• Watering cadence: stable root moisture → stable transpiration → easier VPD control.

Verification loop

1. Set temperature/relative humidity for your stage of growth.
2. Let the room settle; then measure T_leaf and T_air.
3. Compute VPD_leaf and log it (lights on, mid-cycle, pre-lights-off for the first week).
4. Change one variable at a time.

Troubleshooting by symptoms

• Edges tacoing, petioles up, leaves feel dry/dull → VPD or light too intense. Drop VPD by ~0.2–0.3 kPa or reduce PPFD, then re-measure leaf temperature.
• Droopy, soft leaves + wet media → VPD too low and roots suffocating. Raise VPD ~0.2–0.3 kPa; improve airflow; extend time between waterings.
• Powdery mildew pressure → keep late-flower VPD ≥ 1.3 kPa, prune for airflow, sanitize surfaces.

Common mistakes (and easy fixes)

• Using air temperature for both terms. Fix: compute VPDleaf = SVP(Tleaf) – (RH ÷ 100 × SVP(Tair)).
• Single-point measurements. Fix: sample a grid at canopy height.
• Post-watering surprises. Leaves cool more right after irrigation—re-check LTO.
• Chasing decimals. The plant doesn’t care about three decimal places; it cares about stability.

TL;DR workflow you can repeat

• Measure T_air (canopy) and several T_leaf readings; compute LTO.
• Use LTO in your controller or compute VPD with leaf temperature in the left term.
• Hold VPD steady in the stage-appropriate band; only change one input at a time.
• Log weekly: T_air, T_leaf, RH, VPD, and any plant responses.