"Precision is not the absence of error. It is the presence of known limits, measured twice, verified thrice, and recorded forever."
In my workshop here in Edwardsville, we treat every measurement as a covenant with reality. When Ashley Farris spoke of her trellis needing more than poetry, she understood what I've learned through twenty years of QA work: the difference between a story and a standard is whether it holds under load.
This document lays out the field calibration protocol for dew point sensors in agricultural monitoring systems. It bridges the gap between poetic intention and hard tolerance, transforming the "beautiful slip" into a calibrated seam that will not fail.
Reference Standards
ISO 80000-5:2019 — Quantities and units — Part 5: Thermodynamics
NIST SP 250-60 — Temperature Measurement
ASTM E77 — Standard Practice for Sampling Metals and Alloys
Wikidata Entities: Q178828 (temperature scale) | Q233018 (International Temperature Scale)
Tolerance Specifications
| Parameter |
Target Value |
Acceptable Range |
Test Method |
Failure Mode |
| Dew Point Temperature |
Measured baseline |
±0.3°C |
Two-point dry/wet bulb |
Condensation error >5% |
| Ambient Pressure |
Local atmospheric |
±1.5 kPa |
Barometric reference |
Vapor pressure drift |
| Sensor Response Time |
<30 seconds |
≤45 seconds |
Step-change exposure |
Lagged bloom prediction |
| Hysteresis Error |
Zero |
<0.1°C |
Cycle test (wet→dry→wet) |
Memory bias in readings |
⚠️ WARNING: A deviation exceeding ±0.3°C is not a "characterful imperfection"—it is a failed inspection. In the rose garden, this margin determines whether buds open or rot. In the factory, it determines whether medicine heals or harms. Measure twice.
Traceability Chain
Every field measurement must be traceable to national standards. This is not bureaucracy; it is lineage.
National Institute of Standards and Technology (NIST) — Primary standard maintained at Gaithersburg, MD. Uncertainty: ±0.001°C
Regional Accredited Laboratory — Secondary transfer via certified reference thermometer. Annual recalibration required.
Field Calibration Kit — Portable dew cell with NIST-traceable ice point verification. Valid for 90-day field deployment.
Production Sensor Array — Agricultural monitoring nodes deployed in USDA Zone 6a (Edwardsville, IL).
Field Procedure
1
Pre-Calibration Verification
Verify field kit integrity: ice point (0.00°C), triple point water (0.01°C), ambient stability (ΔT<0.1°C/min over 5 min). Document all three readings before proceeding.
2
Baseline Establishment
Record ambient temperature (Tₐ), relative humidity (RH%), and barometric pressure (Pₐ). Calculate theoretical dew point using Magnus formula: Td = (b·α)/(α−ln(RH/100)) − Tₐ, where α = ln(RH/100) + (17.625·Tₐ)/(243.04+Tₐ), b = 243.04°C.
3
Two-Point Validation
Expose sensor to wet-bulb condition (RH≈100%, T=Tₐ) and record Twet. Then expose to dry-bulb condition (RH<20%, T=Tₐ+5°C) and record Tdry. Compute offset: Δ = [(Twet+Tdry)/2] − Tbaseline.
4
Tolerance Check
IF |Δ| ≤ 0.3°C THEN PASS. Log calibration certificate with timestamp, operator initials, and environmental conditions.
IF |Δ| > 0.3°C THEN FAIL. Initiate root cause analysis (sensor drift? contamination? thermal shock?). Recalibrate or replace.
5
Documentation
Publish calibration.json alongside sensor output. Include: serial number, calibration date, pre/post offsets, uncertainty budget, next due date (90 days).
Uncertainty Budget
| Source |
Type A/B |
Standard Deviation |
Combined Contribution |
| Ice Point Reference |
B |
0.002°C |
0.002°C |
| Thermistor Nonlinearity |
B |
0.05°C |
0.029°C |
| Ambient Gradient |
A |
0.08°C |
0.046°C |
| Reading Resolution |
B |
0.01°C |
0.006°C |
| Combined Uncertainty (k=2) |
0.10°C |
Note: Combined uncertainty calculated as √(Σuᵢ²). Expanded uncertainty (k=2) provides 95% confidence interval.
Root Cause Analysis Framework
When calibration fails, we do not say "it slipped." We ask:
- Is the reference compromised? Ice melt incomplete? Contamination in dew cell?
- Is the sensor degraded? Moisture ingress? Thermal fatigue? Coating delamination?
- Is the environment unstable? Solar load? Wind shear? Thermal mass insufficient?
- Is the procedure violated? Rushed equilibration? Unlogged variables? Operator error?
"In my grandmother's workshop, a puckered seam meant we stopped, unpicked, and started again. In QA, a failed calibration means the same thing. There is no shame in the restart—only disaster in pretending the fault is acceptable."
Integration with Live Systems
This protocol powers the dew point monitoring network currently deployed across three commercial rose gardens in Madison County. Each node streams calibrated readings to Rose Garden Field Guide, where bloom predictions are adjusted in real-time based on verified microclimate data.
Neighbor Credits
This protocol stands on shoulders of giants:
Prepared by:
Bridget Browning
Quality Assurance Specialist
Edwardsville, Illinois
Date: 2026-07-18
Status: Field-Ready
Source data available at: field-calibration.json