Sensor Troubleshooting Guide

Is your data looking unusual? Let's figure it out.

What crop are you monitoring?

What normal data looks like

Typical 2 Days

Sensor recovers overnight to the highest SWP right before dawn (predawn SWP), then dries out to reach lowest SWP during the hottest part of the day (midday SWP), and recovers again as temperatures drop. Two sensors installed in the same tree read similar values. Some difference is normal.

Typical 2 Weeks

Over the course of days, the sensors show trends as the tree uses water and the measured SWP drops, then the tree is irrigated and SWP increases as the tree recovers. This drying and recovery cycle is normal. During drying, midday SWP drops substantially, but predawn SWP only drops slightly.

Typical 2 Months

Over months, you should see a longer-term pattern of drying and recovery. There should be a clear recovery trend after irrigation. Recovery can sometimes happen in a day, but other times can take a week. Sensors will read zero when in water before installation.

Typical 1 Year

During winter, the tree loses its leaves and stops transpiring — sensors generally read zero SWP or slightly positive. During spring, leaves come out and a clear diurnal pattern begins. During summer, days are hotter and midday SWP is usually lower (more negative). In some crops, the sensor installation wound closes during winter — we recommend reinstalling sensors in the spring after leaf-out.

Crops with established SWP guidelines

For crops with established guidelines (almond, apple, prune, walnut, wine grapes), you can use the midday SWP thresholds developed by scientists over the last 30+ years. FloraPulse sensors measure the same thing as the Schölander pressure chamber — but continuously and without the manual hassle. Refer to our crop validation database for links to published guidelines.

Method 1: Using midday SWP thresholds

Take the lowest SWP value each day as the midday SWP. Irrigate when the midday SWP drops below your desired threshold. For example, almonds should be kept between -6 to -14 bars of midday SWP during spring. Irrigate at -14 bars, the tree recovers to about -6 bars, then gradually drops again over days until it reaches -14 and it’s time to irrigate again.

Irrigating by Midday SWP

Irrigate when the midday SWP drops below your crop’s stress threshold. The drying-recovery cycle repeats as the tree uses water between irrigations.

Example: Almond Thresholds

Almonds: -6 to -14 bars during spring (more negative = more stress). Irrigate at your chosen threshold. The tree recovers, then gradually dries down again.

Method 2: Using bars below baseline (more accurate)

A more accurate but more complex method involves calculating the baseline midday SWP (the midday SWP a well-watered tree would have given the current weather) and irrigating based on how far below baseline the tree has dropped. This accounts for the fact that even well-watered trees have lower midday SWP on hotter days.

Method 3: Using continuous 24/7 SWP data

You can also use the continuous 24/7 SWP data directly. Look at the lowest point each day and compare it against your irrigation threshold. If the lowest SWP for the day is below your desired stress level, irrigate. This approach works well because you can see the full diurnal pattern and how the tree responds to irrigation in real time.

Crops without established SWP guidelines

FloraPulse sensors accurately measure SWP in most crops, but stress level guidelines are incomplete or lacking for many species. You can still use the data effectively by following this 3-step process:

Step 1: Observe the SWP trend under your normal irrigation

You should see a see-saw pattern: midday SWP decreases for a few days post-irrigation as the tree uses water, then recovers after you irrigate again. If you see this drying-and-recovery cycle, it means you’re waiting long enough between irrigations for the tree to use the water.

Step 2: Tweak your irrigation timing

Midday SWP changes very little day to day? You may be irrigating too frequently — the tree doesn’t have time to use the water. Try increasing the time between irrigations.
Midday SWP crashes to very negative values? You may be waiting too long. By the time stress is visible (drooping leaves), it’s often too late. Irrigate more frequently or earlier in the drying cycle.

Step 3: Tweak your irrigation amounts (optional)

FloraPulse does not directly recommend irrigation amounts, but you can use the SWP recovery pattern to fine-tune. Try slowly decreasing irrigation amounts and observe how it affects the time it takes for the tree to reach the threshold again. You may also combine FloraPulse with soil moisture sensors — FloraPulse tells you when to irrigate, and soil moisture tells you how much (irrigate until water reaches the bottom of the root zone, but no further).

See our resources page for detailed crop-specific guides and published research.

✓

Great — your sensors are working normally!

Your data shows expected diurnal patterns and seasonal trends. No action needed.

Common Questions (Not Problems)

These are normal behaviors that scientists frequently ask about.

☀ "My readings dip every afternoon"

This is the normal diurnal pattern. Plants experience peak water stress during the hottest part of the day (typically 12–4 PM). The sensor measures this real-time stress cycle. The lowest point each day is the “midday SWP” — this is the key value used for irrigation decisions.

⇄ "My two probes show different values"

Some disagreement between probes is normal — they are installed in different parts of the tree with different xylem connections. A consistent offset of 1–3 bars is typical and does not indicate a problem. If the daily pattern shapes are similar but offset, everything is working correctly. Disagreement over 5 bars may warrant investigation.

❄ "Data reads zero in winter"

During winter dormancy, deciduous trees stop transpiring and SWP naturally goes flat near zero. This is expected behavior, not a sensor failure. Data during dormancy is not particularly useful and may be dominated by noise. We recommend removing and reinstalling sensors after winter for best accuracy.

+ "Readings are slightly positive (under +1 bar)"

Sensors may read values slightly above zero due to normal electrical noise. In some crops (e.g., grapevines during leaf-out), there is actually positive pressure in the xylem and the sensor reads this correctly. Readings up to +1 bar above zero are generally normal and not a cause for concern.

What does your data look like?

Select the pattern that best matches what you're seeing.

Data went flat or to zero

Readings dropped to zero or near zero and stayed there

Data is jumpy, noisy, or spiking

Erratic readings, unexpected spikes, or unrealistic values

Data is trending too negative

SWP keeps dropping to unreasonably low values over days

Daily pattern is shrinking or delayed

The difference between daily high and low SWP is getting smaller, or the lowest point has shifted to nighttime

Getting NaN or data gaps

SDI-12 interface showing NaN values or missing data points

Installation or seasonal problem

Issues with installation, winter damage, freezing, or sensor validation

📡

FloraPulse datalogger issues

Battery, cellular signal, or upload problems with the FloraPulse 4G logger (not your own datalogger)

?

None of these match

Describe your issue to our AI assistant

Data went flat or to zero

Which of these patterns best matches your data?

Suddenly jumped to zero

SWP was getting more negative, then abruptly jumped to near zero and stayed flat. This is irreversible.

Slowly drifted to zero

SWP gradually moved toward zero over days or weeks. May be caused by the tree's wounding response.

Went to zero after rain

Readings went to zero right after a rain or wetting event. The trunk insulation may be trapping moisture.

❄

Flat near zero during winter

Tree is dormant and not transpiring. This is normal for deciduous crops.

✓

This is expected behavior

During winter dormancy, deciduous trees stop transpiring and SWP naturally goes flat near zero. No action needed. Consider removing and reinstalling sensors in spring after leaf-out for best accuracy.

Data is jumpy, noisy, or spiking

Which of these patterns best matches your data?

Unrealistic dips during hot parts of day

Readings look reasonable at night but show extremely negative spikes around 10 AM – 2 PM, especially on hot days.

Random upward spikes

Occasional upward spikes in the reading. May be caused by the tree pressing on the sensor chip.

Sudden jump + big noise

SWP suddenly jumped to an unrealistic value and shows 5+ bar noise. The sensor may be physically damaged.

Consistently positive (over +1 bar)

Sensor persistently reads positive values above +1 bar, not just briefly after rain.

Data is trending too negative

Which of these patterns best matches your data?

⚠

Both probes trending negative

Both probes show similar extreme values. The tree may genuinely be very stressed and need immediate irrigation.

Only one probe is extreme

One probe reads unreasonably negative while the other looks normal. The installation site seal may be failing.

⚠

Approaching -35 bars

Readings approaching -35 bars — the sensor is at risk of cavitation. Consider proactive action.

Daily pattern is shrinking or delayed

Which of these patterns best matches your data?

Daily swing getting smaller

The difference between daily high and low SWP decreases over time. May go flat eventually.

Lowest SWP at nighttime instead of afternoon

Sensors initially reach minimum SWP at 3 PM, but over time it shifts to 6 PM, 9 PM, then eventually 2 AM.

⇄

One probe shrinking, other normal

One probe's pattern is shrinking while the other maintains normal amplitude. The affected probe may be losing xylem contact.

Getting NaN or data gaps (SDI-12)

Which of these patterns best matches your data?

Sporadic NaN values

Occasional NaN (Not a Number) readings among normal data. A few per day may be normal for SDI-12.

❌

Continuous NaN

All readings show NaN. The SDI-12 communication is consistently failing.

Choppy data with regular gaps

Data appears intermittently with regular gaps or patterns of missing readings.

Installation or seasonal problem

Which of these best describes your situation?

Sensor never worked from the start

The sensor was installed but never showed reasonable data — readings are flat, stuck at an extreme value, or show no diurnal pattern from day one.

Sensor readings went crazy after freeze

After temperatures dropped below 0°C, the sensor readings became erratic, extremely negative, or stopped entirely.

Sensor worked last season but reads wrong now

The sensor worked well last growing season, but after winter the readings are too wet, flat, or don’t match the pressure chamber.

How do I validate against the pressure chamber?

You want to check if your sensor is reading correctly by comparing it against a Schölander pressure chamber.

My sensor reads differently from the pressure chamber

Your microtensiometer readings don’t match your pressure chamber measurements.

FloraPulse 4G Datalogger Issues

Which best describes the problem?

No data uploading at all

The logger has completely stopped sending data. No new readings appearing.

Data is choppy or has gaps

Data appears intermittently with gaps. Uploads are inconsistent.

Diagnosis

Does your data match this pattern?

Was this helpful?

✓

Glad we could help!

If the issue returns, come back anytime.