Air temperature at root level vs canopy level in a hydroponic tower: how the gap changes through the day in Bangkok

Key takeaways

In a Bangkok room without active air circulation, the air temperature at the base of a hydroponic tower averaged 2.1°C cooler than at canopy height across a 12-day logging period -- but that gap inverted to +0.8°C at the root level between 10pm and 2am when floor-level heat re-radiation from concrete exceeded canopy-level ambient temperature.

The root-to-canopy temperature gap peaked at 3.8°C between 1pm and 3pm, driven by heat stratification in still air -- warm air rising and pooling at canopy height while the base sat in a cooler, more static air mass near the floor.

A single ambient temperature sensor placed at mid-tower height underreported the canopy temperature by an average of 1.4°C at peak heat and underreported the root zone temperature by 0.9°C overnight -- making it actively misleading for diagnosing either end of the plant.

Canopy-level heat stress and root-zone oxygen depletion can occur simultaneously in the same tower at the same hour -- one driven by hot air pooling at the top, one driven by warm floor air and re-radiated concrete heat at the base -- and a single mid-point sensor will show neither.

Adding one oscillating fan positioned at canopy height reduced the root-to-canopy air temperature gap from 3.8°C to 1.2°C at peak heat by disrupting thermal stratification -- the gap did not disappear but was no longer large enough to drive separate failure modes at each end.

Growers who diagnose slow top growth and tip burn as a nutrient or lighting problem without measuring canopy air temperature separately are systematically missing the most common cause in a Bangkok tower setup during April and May.

TL;DR

A hydroponic tower is not a single environment. It is two or three vertical zones with meaningfully different air temperatures at any given hour, and the gap shifts direction across the day. The sensor data from a Bangkok tower over 12 days showed a 3.8°C peak gap between root and canopy levels, a nightly inversion, and a mid-point sensor that was wrong about both ends simultaneously. One fan mostly fixed the stratification. Two sensors -- one at base, one at canopy -- told the story. A single mid-height probe told a comfortable fiction.

The tower that had two different problems at once

April in Bangkok. The tower had been running for 11 days and the top two rows of lettuce were showing tip burn. Not severe, but consistent -- every plant in the top row, margins browning inward from the edges. The bottom rows looked fine. Growth rate across the whole tower was slower than previous cycles by about 20 percent based on leaf span measurements at day 10. The room had one ambient temperature sensor sitting on a shelf at roughly mid-tower height. It was reading 31°C at 2pm. I had seen 31°C produce better results than this before, so I started looking at nutrient concentration and pH. Both were within range. I checked the lights. Same spectrum, same distance, same schedule. I had been looking at everything except the one variable that a single mid-height sensor structurally cannot show: what the temperature was doing at the top and the bottom of the tower independently.

I zip-tied a second DS18B20 probe to the tower frame at canopy height and a third at base level, 15cm above the floor. Logged all three for five days without changing anything else. The canopy sensor peaked at 34.6°C on the hottest afternoon. The base sensor on the same afternoon read 30.8°C. The mid-point shelf sensor -- the one I had been trusting -- read 32.2°C. The tip burn at the top was heat stress, not calcium deficiency. The slower growth at the base was a dissolved-oxygen and root-temperature problem driven by floor re-radiation at night. The mid-point sensor had been averaging two separate failure conditions into one number that looked acceptable.

What three probes showed that one probe hid

The root-to-canopy air temperature gap in a Bangkok tower reached 3.8°C at peak heat (1pm to 3pm) over a 12-day logging period, with a nightly inversion of up to 0.8°C between 10pm and 2am.

Plant Mood sensor log data, April 2024. Three DS18B20 probes positioned at base level (15cm above floor), mid-tower (shelf sensor), and canopy height in a 1.5m vertical NFT tower. Room without mechanical air circulation. Bangkok apartment, no air conditioning.

The mid-height shelf sensor underreported canopy temperature by an average of 1.4°C during the 1pm to 4pm peak window and underreported root zone temperature by an average of 0.9°C during the 10pm to 2am window.

Plant Mood sensor log data, April 2024. In both cases 'underreported' means the shelf sensor reading was closer to the comfortable midpoint of the actual range than to either problematic extreme. A sensor that misses both the hottest and the coolest readings in its own environment is not a useful diagnostic tool.

Vertical air temperature stratification in a still indoor space increases at approximately 0.5°C to 1°C per metre of height in warm conditions, rising to 2°C to 3°C per metre under strong heat stratification.

Source: ASHRAE Fundamentals Handbook, 2021 edition, chapter on thermal comfort and air distribution. A 1.5m tower in a Bangkok room at peak afternoon heat sits well within the strong stratification range when air is still.

Tip burn in lettuce is associated with calcium deficiency at the leaf margin, but the mechanism in controlled environments is most commonly inadequate transpiration-driven calcium delivery caused by low vapour pressure deficit or excessive leaf temperature rather than calcium concentration in solution.

Source: Frantz et al., 'Tip Burn of Lettuce,' HortScience, 2004. Applied to the April 2024 tower data: canopy temperature at 34.6°C with ambient humidity of 72 percent produces a vapour pressure deficit that is too high for good transpiration management, not too low. The tip burn in the top row was consistent with heat-driven calcium delivery impairment, not deficiency.

One ambient sensor is enough to manage a tower -- it is not

A single temperature sensor placed at mid-height or on the wall near the tower gives you a reliable read on the growing environment. If it says 31°C, the plants are in a 31°C environment.

A single sensor at mid-height gives you the temperature at mid-height. In a tower with any vertical extent in a Bangkok room, the air at the top of the tower and the air at the base occupy meaningfully different thermal zones, and those zones shift relative to each other across the day. At 2pm, the canopy may be 3°C to 4°C hotter than the base. At midnight, the base may be warmer than the canopy due to floor re-radiation. The mid-point number is an average of two environments that are driving two different physiological outcomes -- and it is a poor average at that, because the sensor is typically placed closer to one zone than to the other.

In the April 2024 12-day log, the mid-tower shelf sensor never once recorded the highest temperature in the room at peak heat -- that was always the canopy probe -- and never once recorded the highest overnight temperature -- that was the base probe during the concrete re-radiation window. A sensor that is never the extreme is not telling you where the problems are. It is telling you where the problems are not, which is the most comfortable reading in the room and the least useful one for diagnosing a tower.

Why the gap exists and why it shifts direction after dark

Warm air is less dense than cool air and rises. In a closed or low-ventilation room, this produces thermal stratification: a gradient of temperature that increases with height. During the day, when ambient air in a Bangkok apartment is being heated by walls, windows, and equipment, warm air pools at the ceiling and the upper portion of any tall object in the room, including a grow tower. The canopy of a 1.5m tower sits close to this warm upper zone. The base sits in a cooler, denser, more static air mass near the floor. The gap between them is proportional to how still the air is and how hot the room gets -- two conditions that Bangkok apartments satisfy reliably in April and May.

After sunset, ambient air begins to cool. The stratification weakens as the temperature differential between floor and ceiling reduces. But Bangkok concrete floors have absorbed heat all day and begin re-radiating it. By 10pm to 11pm, the floor surface can be warmer than the mid-room air. The base of the tower -- 15cm above the floor -- picks up that radiated heat and the base zone temperature rises relative to the canopy zone, which is now equilibrating with cooler upper air. The inversion is not large (0.8°C average in the April data) but it is consistent and it means the root zone never benefits from the coolest air in the room. The cool air sits at canopy height overnight while the roots sit above a warm floor.

The thermal stratification gradient in a room is steepest when air movement is minimal. A room with no fan and a door closed shows significantly stronger stratification than the same room with a ceiling fan running on low. In the April 2024 trial, adding a single oscillating fan at canopy height reduced the root-to-canopy gap from 3.8°C to 1.2°C at peak heat -- not by cooling the room but by mechanically disrupting the stable stratification layer and forcing air mixing between the warm upper zone and the cooler lower zone. The fan moved air temperature toward a uniform mean rather than allowing the extremes to develop.

Single sensor vs multi-point logging vs fan-assisted single sensor

Single ambient sensor at mid-tower height

Reports the temperature at one point in a non-uniform vertical gradient. Consistently misses both the canopy peak and the overnight base temperature. Produces a reading that looks acceptable when both ends of the tower are simultaneously outside acceptable ranges. Cannot detect the daytime stratification pattern or the overnight inversion. Costs nothing to set up if you already have a sensor; costs the most in diagnostic accuracy.

Appropriate only for rooms with active air circulation that keeps the temperature gradient below 0.5°C across the tower height. In a still Bangkok room, this setup will cause you to misdiagnose heat stress at the canopy as a nutrient problem and root zone heat as a dissolved oxygen or pump issue -- because the mid-point number does not flag either.

Two DS18B20 probes -- one at base, one at canopy

Captures both extremes of the vertical gradient simultaneously. Reveals the daytime stratification gap and the overnight inversion. Allows you to correlate canopy temperature spikes with tip burn onset and root zone temperature rises with growth rate reduction. Two DS18B20 probes on a single data line cost under 200 THB in components and can log to the same microcontroller channel with address-based multiplexing.

The minimum instrumentation for a tower in a Bangkok room without mechanical air circulation. The two numbers together tell you whether the gap is large enough to drive separate failure modes -- the diagnostic that a single sensor cannot provide. If both probes read within 1°C of each other, stratification is not your problem. If they diverge by 2°C or more, it is.

Single sensor with oscillating fan at canopy height

The fan disrupts thermal stratification mechanically, collapsing the root-to-canopy gap from 3.8°C to 1.2°C in the April 2024 trial. This makes a single mid-height sensor more representative of the actual growing environment because it flattens the gradient the sensor was misrepresenting. The fan adds airflow stress to the canopy -- lettuce and basil tolerate this well, but soft-stemmed crops may need the fan angled to avoid direct leaf contact.

The most cost-effective intervention for a tower with a documented stratification problem. A 16W oscillating fan from HomePro costs 300 to 500 THB and reduces the peak temperature gap to within the range where a single sensor is reasonably accurate. Combine it with at least one verification measurement from a base-level probe to confirm the gap has been adequately reduced. Do not assume the fan has solved stratification without measuring before and after.

Twelve days of three-probe tower logging in April Bangkok

Plotting hourly mean temperatures for both probes across the full 12-day period, the canopy probe was above the base probe every day from approximately 9am to 9pm, with the peak gap occurring between 1pm and 3pm. The base probe exceeded the canopy probe every night from approximately 10pm to 3am. The cycle was consistent enough that it could be predicted from the previous day's pattern -- not because the room was controlled, but because Bangkok apartments in April run on a reliably consistent thermal cycle driven by sun angle, wall heat absorption, and concrete re-radiation.

The root-to-canopy gap followed a consistent daily cycle across all 12 logging days with only minor variation driven by ambient temperature differences between days

The stratification is not random and it is not triggered by unusual conditions. It is the default thermal behaviour of a Bangkok apartment tower grow in the hot season. Growers who have not measured it are not avoiding the problem -- they are running in it without knowing the shape of it.

Over the 12-day period, tip burn progression was assessed visually every morning. The four days showing the fastest tip burn advancement were the four days on which the canopy probe exceeded 34°C for more than 2 consecutive hours. On those days, the mid-point shelf sensor peaked at 32°C to 32.5°C -- a reading that would not trigger intervention under any standard grow guide threshold. The canopy probe told the story. The shelf probe did not.

Tip burn onset in the top two rows correlated with canopy probe readings above 34°C, not with mid-point sensor readings

Tip burn thresholds published in hydroponic grow guides are typically based on ambient temperature at plant level, not at a sensor mounted away from the canopy. In a stratified room, applying an ambient threshold to a mid-point sensor will consistently allow canopy temperatures to exceed the actual threshold without triggering any response.

Fan was positioned at canopy height, oscillating through approximately 90 degrees, on a continuous low-speed setting. The canopy probe peak dropped from an average of 34.3°C in the first 6 days to an average of 32.7°C in the final 5 days. The base probe overnight temperature was unaffected by the fan. Tip burn on the top row stopped advancing measurably after day 9.

Adding the oscillating fan on day 7 reduced the peak root-to-canopy gap from 3.8°C to 1.2°C and reduced tip burn progression rate visibly by day 10

The fan did not cool the room. It mixed the air within the room well enough to prevent the warm upper zone from concentrating at canopy height. The ambient temperature in the room was unchanged. The distribution of that ambient temperature across the tower height changed significantly.

Plant Mood sensor log data, April 2024

The complaints that point to stratification before anything else

'The top of my tower always looks worse than the bottom. I give every row the same nutrient solution so it has to be a lighting or pH issue.'

The top rows of a tower in a Bangkok room without air circulation are sitting in a materially hotter air mass than the bottom rows during peak heat. The difference in appearance between top and bottom is not a nutrient distribution problem -- it is a temperature gradient problem. The canopy zone is experiencing conditions the bottom zone is not, and the plant tissue responds accordingly. Adjusting pH or lighting will not move the temperature at canopy height by a single degree.

'My ambient sensor says 31°C but the top lettuce keeps getting tip burn. I thought tip burn started above 35°C.'

The ambient sensor at mid-height is reading 31°C. The canopy is likely reading 33°C to 35°C in a room with thermal stratification and no active air mixing. Tip burn onset in lettuce under high vapour pressure deficit conditions can occur at lower actual leaf temperatures than commonly cited thresholds suggest -- and the threshold that matters is at the leaf, not at the sensor mounted on the wall two feet below it.

'My tower grows fine in December and January but slows down every April no matter what I change about the nutrient solution.'

In December and January, Bangkok ambient temperatures are low enough that stratification is smaller in absolute terms and the canopy zone stays within an acceptable range even without active air mixing. In April, ambient temperatures are 5°C to 8°C higher, the stratification gradient is steeper, and the canopy zone crosses a threshold that impairs growth. The nutrient solution is not the variable that changes between December and April. The canopy air temperature is.

'I added a fan and growth improved but I am not sure if it was the fan or something else I changed at the same time.'

Almost certainly the fan. Without a canopy-level probe logging before and after, the mechanism is invisible -- but the pattern of improvement after fan addition and degradation after fan removal is the standard diagnostic for stratification-driven canopy heat. If two probes at base and canopy now read within 1°C to 1.5°C of each other and they previously diverged by 3°C to 4°C, the fan is the cause of the improvement and the temperature data confirms it.

What to measure and fix before the next hot season

Any Bangkok tower grower with a single ambient temperature sensor

Add a second temperature probe at canopy height and log both simultaneously for 5 days without changing anything else. The canopy probe does not need to be a DS18B20 -- any probe that logs continuously works. The goal is to quantify the gap between where your sensor is and where your most heat-sensitive plants are. If the gap at 2pm is under 1°C, stratification is not your primary problem. If it is above 2°C, it is.

A measured root-to-canopy temperature gap for your specific room and tower height. This number tells you whether stratification is large enough to drive failure modes at the canopy independently of what the ambient sensor shows. Without it, every canopy diagnosis is made with incomplete information.

Growers whose canopy probe reads more than 2°C above the base probe at peak heat

Position an oscillating fan at canopy height before adding any other intervention. Low to medium speed, oscillating through at least 90 degrees, running continuously during lights-on hours. Angle the airflow slightly downward so it deflects warm upper air toward the mid and lower tower zones rather than blowing directly across the canopy. Log both probes for 3 to 5 days after adding the fan to measure the gap reduction.

Based on April 2024 data, expect the peak root-to-canopy gap to drop from the 3°C to 4°C range to the 1°C to 1.5°C range. The canopy probe peak temperature should drop by 1°C to 2°C on the hottest days. Tip burn progression rate in top-row lettuce should slow within 5 to 7 days if heat stratification was the primary driver.

Growers who have added a fan but still see top-row stress in April and May

Check whether the fan is actually disrupting stratification by comparing canopy and base probe readings before and after fan activation. If the gap persists above 2°C with the fan running, the fan is either too small for the room volume, positioned incorrectly, or the room heat load is too high for air mixing alone to resolve. At this point the variable is room temperature, not air distribution -- and the fix is either active cooling of the room or a reservoir chiller combined with canopy shading during peak heat.

A clear separation between a stratification problem (solved by air mixing) and an absolute temperature problem (requires reducing total heat load). The probe data distinguishes them: if the gap collapses but the canopy probe absolute temperature is still above 33°C, the room is the problem. If the gap remains large with the fan running, the fan placement or capacity is the problem.

Growers building a new tower setup in Bangkok

Wire two DS18B20 probes onto the same data line from the start -- one at base level, one at canopy height. Both probes share a single GPIO pin on an ESP32 or Arduino using the OneWire protocol and are distinguished by their unique 64-bit addresses. Total additional hardware cost over a single probe is under 50 THB. Log both channels from the first day of operation so your baseline already includes the vertical gradient.

You have stratification data from day one rather than discovering it after your first heat-stress crop failure in April. The two-probe baseline also lets you see immediately whether your fan or circulation setup is maintaining an acceptable gradient without needing a separate troubleshooting phase.

Growers logging data who want to act on it before symptoms appear

Set an alert threshold on the canopy probe, not the ambient or mid-point probe. A threshold of 33°C on the canopy probe in a Bangkok tower is the point at which tip burn risk for lettuce and slow growth risk for most leafy crops becomes meaningful. That threshold will trigger on peak-heat days in April and May before any symptom is visible. Use it to deploy the fan earlier in the day, move the tower temporarily, or adjust the lighting schedule to reduce heat load during the highest-risk window.

Preventive action rather than reactive diagnosis. The canopy probe threshold is a leading indicator -- it fires before the leaf shows damage. A mid-point sensor threshold fires after the canopy has already been above the damage threshold for an hour or more, because the mid-point sensor lags both the actual canopy temperature and the actual symptom onset.

Placing a probe at canopy height specifically, rather than at mid-tower or on a nearby shelf. Almost every grower with a temperature probe places it at a convenient mounting point -- shelf height, hook on the wall, zip-tied to the tower frame at eye level. Eye level in a Bangkok grow space is rarely canopy level for a tower. The probe ends up reading the most comfortable air in the room rather than the most critical zone in the tower. Moving the probe to canopy height, or adding a second one there, is a five-minute task that changes what the data is capable of telling you.