Combined standard uncertainty is the standard uncertainty assigned to a result calculated from several input quantities. It combines the uncertainty contributions of the inputs while accounting for their sensitivity coefficients and, when necessary, correlations.
Combined Standard Uncertainty
Combined standard uncertainty is the standard uncertainty assigned to a result calculated from several input quantities. It combines the uncertainty contributions of the inputs while accounting for their sensitivity coefficients and, when necessary, correlations.
\(u_{\mathrm{c}}^{2}(y) = \sum_{i=1}^{n} [\partial f/\partial x_{i}]^{2}u^{2}(x_{i}) + 2\sum_{i<j}(\partial f/\partial x_{i})(\partial f/\partial x_{j})cov(x_{i},x_{j})\)
What this formula means
Scientific principle
Principle
A calculated result depends on input quantities through a mathematical model. Each input contributes uncertainty according to how strongly the result changes when that input changes. Independent contributions combine through variance addition, while correlated inputs introduce covariance terms.
Derivation
1. Let the calculated result be y = f(x₁, x₂, ..., xₙ).
2. For small input variations, approximate the change in y using a first-order Taylor expansion.
3. The sensitivity coefficient for input xᵢ is cᵢ = ∂f/∂xᵢ.
4. Each input contributes a variance term cᵢ²u²(xᵢ).
5. If inputs are correlated, each pair contributes a covariance term 2cᵢcⱼcov(xᵢ,xⱼ).
6. Add all variance and covariance contributions.
7. The square root of the resulting variance gives the combined standard uncertainty.
Final Result
u_c(y) = √{Σᵢ₌₁ⁿ cᵢ²u²(xᵢ) + 2Σᵢ<ⱼ cᵢcⱼcov(xᵢ,xⱼ)}
Independent Input Relation
u_c(y) = √[Σᵢ₌₁ⁿ cᵢ²u²(xᵢ)]
Sensitivity Coefficient
cᵢ = ∂f/∂xᵢ
Covariance Relation
cov(xᵢ,xⱼ) = rᵢⱼu(xᵢ)u(xⱼ)
Variables and their meanings
| Symbol | Meaning | Unit |
|---|---|---|
y | Calculated measurement result | Unit determined by the mathematical model |
xᵢ | ith input quantity | Unit of the input quantity |
u(xᵢ) | Standard uncertainty of input xᵢ | Same unit as xᵢ |
u_c(y) | Combined standard uncertainty of the calculated result | Same unit as y |
cᵢ | Sensitivity coefficient for input xᵢ | Unit of y divided by unit of xᵢ |
cov(xᵢ,xⱼ) | Covariance between two input quantities | Product of their units |
rᵢⱼ | Correlation coefficient between two inputs | Dimensionless, between −1 and +1 |
SI units and conversions
Combined Standard Uncertainty
Same unit as calculated result
Variance
Square of the unit of calculated result
Sensitivity Coefficient
Output unit divided by input unit
Correlation Coefficient
Dimensionless
Conditions of use
The result must be expressed as a sufficiently differentiable function of its input quantities.
The uncertainty propagation is based on a first-order Taylor approximation.
Input standard uncertainties must be estimated using appropriate methods.
Covariance terms must be included when input quantities are correlated.
The uncertainty contributions must be expressed using consistent units.
Limitations and assumptions
First-order propagation may be inadequate for highly nonlinear models or large uncertainties.
Ignoring significant correlations can produce an incorrect combined uncertainty.
The method does not automatically account for model inadequacy or unrecognized systematic effects.
A calculated standard uncertainty is not itself a guaranteed confidence interval.
More advanced methods, such as Monte Carlo propagation, may be appropriate for strongly nonlinear models or non-Gaussian input distributions.
Step-by-step worked examples
| Problem | Solution | Answer |
|---|---|---|
A calculated quantity is y = x₁ + 2x₂. The independent input standard uncertainties are u(x₁) = 0.3 m and u(x₂) = 0.4 m. Calculate the combined standard uncertainty. |
| The combined standard uncertainty is approximately 0.854 m. |
A result is calculated using y = x₁ + x₂. Both inputs have standard uncertainty 0.5 units and a correlation coefficient of 0.6. Calculate the combined standard uncertainty. |
| The combined standard uncertainty is approximately 0.894 units. |
Real-world applications
Scientific measurement uncertainty budgets
Calibration laboratory calculations
Engineering measurement systems
Uncertainty analysis of derived physical quantities
Metrology and experimental research
Formula Meaning
Combined standard uncertainty is the standard uncertainty assigned to a result calculated from several input quantities. It combines the uncertainty contributions of the inputs while accounting for their sensitivity coefficients and, when necessary, correlations.
Sources and further reading
No source has been supplied for this item yet. Treat the claim as needing editorial verification.