Why greenhouse stress can exceed 24 hours

The greenhouse comparison reports 29.9 stress axis-hours per day during the planner-offline window. That number can look impossible if it is read as elapsed clock time. It is not saying that a day lasted almost 30 hours.

The metric evaluates two different environmental axes during every clock hour:

• Temperature: Was the greenhouse temperature outside its target range?
• Vapor pressure deficit (VPD): Was VPD outside its target range?

Each axis has 24 possible hours in a day. Together, they create 48 possible axis-hours per day: 24 temperature axis-hours plus 24 VPD axis-hours.

How an hour is counted

One clock hour can contribute:

• 0 axis-hours when both temperature and VPD are inside their target ranges;
• 1 axis-hour when either temperature or VPD is outside its range; or
• 2 axis-hours when temperature and VPD are both outside their ranges.

That last case is why the total can exceed 24. The two dimensions are counted separately, even when their stress occurs during the same clock hour.

How to read the comparison

The planner-offline window averaged 29.9 out of 48 possible stress axis-hours per day. The AI-planning-online window averaged 13.1 out of 48. Lower is better for this metric.

Those values summarize the combined duration of temperature and VPD stress. They do not say that plants experienced 29.9 distinct elapsed hours of stress, and they do not establish that AI planning alone caused the difference between the two operating windows.

The canonical fixed-window comparison publishes the underlying context and caveats. The greenhouse case study explains how Verdify connects those measurements to a governed AI learning loop while keeping weather, equipment, crop, and instrumentation limits visible.