Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
ISO/TS 25377:2007 — Hydrometric Uncertainty Guidance (HUG)
Measurement uncertainty in hydrometry — Application of GUM to flow measurement in open channels
Introduction to Hydrometric Uncertainty Guidance
The management of natural water resources and the environment fundamentally depends on the quality of hydrological measurements. Without a reliable understanding of measurement quality, effective water management — from flood protection to irrigation scheduling — becomes guesswork. ISO/TS 25377:2007, known as the Hydrometric Uncertainty Guidance (HUG), fills a critical gap by providing hydrometry-specific guidance on applying the internationally recognized ISO/IEC Guide 98 (GUM) framework.
The essential purpose of the GUM is that a statement of the quality of a measurement result should accompany every measurement described in technical standards. While the GUM serves the universal requirements of metrology, HUG is specifically tailored to hydrometry — the measurement of components of the hydrological cycle. It selects the most applicable methods from the GUM and applies them to the techniques and equipment used in hydrometry.
Before HUG, error analysis was commonly used in hydrometry, but such statements presuppose knowledge of a “true” value — which can never be known. The uncertainty approach properly acknowledges this limitation by focusing on the dispersion of measured values around their mean.
Core Uncertainty Framework and Key Definitions
Standard Uncertainty and Probability Models
The GUM defines standard uncertainty of a result as equivalent to a standard deviation — either the standard deviation of a set of measured values or of probable values derived from probability distributions. HUG emphasizes that this is broadly similar to the error analysis approach but provides additional methods of estimating uncertainty based on probability models when direct measurement data is scarce.
| Uncertainty Type | Source | Evaluation Method | Hydrometry Example |
|---|---|---|---|
| Type A | Statistical analysis of repeated observations | Standard deviation of the mean | Repeated current-meter measurements at a single vertical |
| Type B | Other means (manufacturer specs, prior data, experience) | Assumed probability distributions | Manufacturer-stated accuracy of a stage sensor (±0.1%) |
| Combined | Multiple input quantities | Law of propagation of uncertainty | Flow Q = f(A, V) combining area and velocity uncertainties |
Law of Propagation of Uncertainty
When a measurement result Y is determined from N input quantities X₁, X₂, …, XN, the combined variance is given by:
uc²(y) = Σ(∂f/∂xᵢ)² · u²(xᵢ) + 2 · ΣΣ(∂f/∂xᵢ)(∂f/∂xⱼ) · u(xᵢ, xⱼ)
This elegantly handles both independent and correlated input uncertainties — a crucial capability for hydrometric measurements where variables such as channel width, depth, and velocity are rarely truly independent.
Uncertainty in Open Channel Flow Measurement
Velocity-Area Method
The velocity-area method is the most widely used technique for open channel discharge measurement. HUG provides detailed guidance on evaluating uncertainties in:
Mean velocity determination: The standard deviation of point velocity measurements within a vertical section, considering both the number of measurement points and the integration method used. For a standard 0.2–0.6 depth method (two-point measurement), the relative uncertainty in mean velocity typically ranges from 5% to 15% depending on flow conditions.
Velocity-area integration: The uncertainty arising from spatial integration across the channel cross-section. The number of verticals and the method of interpolation between them significantly affect the overall uncertainty budget.
Perimeter flow: Near-bank and boundary regions contribute disproportionately to uncertainty because velocity gradients are steepest there. HUG provides specific guidance for estimating this often-overlooked component.
Warning: The uncertainty contribution from perimeter (boundary) flow is routinely underestimated in field practice. In narrow or shallow channels, it can account for 30–50% of total discharge uncertainty.
Critical Depth and Dilution Methods
For critical depth structures (flumes, weirs), HUG addresses uncertainty in head measurement, geometry determination, and the iterative calculation of discharge coefficient. For dilution methods (chemical gauging), both continuous feed and transient mass (integration) approaches are covered, with detailed guidance on tracer concentration measurement uncertainty.
Practical Application and Engineering Insights
Monte Carlo Simulation for Complex Uncertainty
HUG introduces Monte Carlo Simulation (MCS) as a powerful tool for evaluating uncertainty in complex hydrometric systems where the law of propagation becomes mathematically intractable. By randomly sampling from input probability distributions thousands of times, MCS directly generates the probability distribution of the output quantity — without requiring linearization or normality assumptions.
Engineering insight: MCS is particularly valuable for rating curve uncertainty analysis, where the relationship between stage and discharge is nonlinear and the parameters are correlated. A typical implementation with 10,000 iterations can reveal confidence intervals that would be impossible to derive analytically.
Equipment Performance Specifications
HUG provides a performance guide for hydrometric equipment used in technical standard examples, covering current meters, acoustic Doppler velocimeters (ADVs), pressure transducers, and radar level sensors. The key recommendation is that manufacturer-stated accuracy should always be verified through field calibration — particularly for instruments deployed in harsh or sediment-laden waters.
Never rely solely on manufacturer uncertainty specifications. Field conditions — biofouling, sediment abrasion, temperature extremes — can degrade instrument performance by a factor of 2–5 compared to laboratory calibration.
Frequently Asked Questions
Q: What is the difference between error analysis and uncertainty analysis in hydrometry?
A: Error analysis assumes a known true value and evaluates deviations from it. Uncertainty analysis acknowledges that the true value cannot be known and instead characterizes the dispersion of values that can be reasonably attributed to the measurand — a more scientifically rigorous approach.
A: Error analysis assumes a known true value and evaluates deviations from it. Uncertainty analysis acknowledges that the true value cannot be known and instead characterizes the dispersion of values that can be reasonably attributed to the measurand — a more scientifically rigorous approach.
Q: How many uncertainty components should I consider for a velocity-area discharge measurement?
A: At minimum: (1) velocity measurement at each vertical, (2) depth measurement, (3) width measurement, (4) number of verticals, (5) integration method, and (6) perimeter flow. In practice, 8–12 components are typical for a thorough analysis.
A: At minimum: (1) velocity measurement at each vertical, (2) depth measurement, (3) width measurement, (4) number of verticals, (5) integration method, and (6) perimeter flow. In practice, 8–12 components are typical for a thorough analysis.
Q: Can HUG be applied to groundwater flow measurements?
A: While HUG focuses on surface water hydrometry (open channels, rivers, pipes), the same GUM-based principles can be extended to groundwater applications with appropriate domain-specific modifications to the uncertainty models.
A: While HUG focuses on surface water hydrometry (open channels, rivers, pipes), the same GUM-based principles can be extended to groundwater applications with appropriate domain-specific modifications to the uncertainty models.
Q: How often should hydrometric instruments be calibrated to maintain valid uncertainty statements?
A: HUG recommends calibration intervals based on instrument type and operating conditions. For permanent gauging stations, annual calibration is a minimum; for instruments in abrasive or biofouling environments, semi-annual or quarterly calibration may be necessary.
A: HUG recommends calibration intervals based on instrument type and operating conditions. For permanent gauging stations, annual calibration is a minimum; for instruments in abrasive or biofouling environments, semi-annual or quarterly calibration may be necessary.
