Uncertainty background and theory

Being able to account for uncertainty is an essential piece of effective carbon accounting. With our quantitative uncertainty accounting system, we give you the ability to easily track and measure the certainty of your inventory data.

Please read on for additional details about the background and theory underpinning our uncertainty functionality and calculations.

Understanding Uncertainty in Carbon Accounting

Unlike financial accounting, carbon accounting is filled with uncertainties due to assumptions about activity data, emission factors, and methodologies. Recognizing and quantifying these uncertainties is vital to interpreting carbon footprint results meaningfully.

Qualitative vs. Quantitative Uncertainty

Traditionally, uncertainty has been assessed qualitatively (e.g., 'high', 'medium', or 'low'). While easier to implement, this method is often vague and difficult to communicate clearly with end-users.

By contrast, quantitative uncertainty provides numerical confidence intervals—such as +/- 40%—that are easier to interpret and defend. It also allows you to identify weak data points and drive improvements.

The Greenhouse Gas Protocol’s Perspective

The GHGP outlines standards for uncertainty management:

• Product Standard: Quantitative uncertainty is required.
• Scope 3 Standard: Quantitative assessment is recommended.

However, the methodology is complex and data-intensive. Many organizations find it impractical to implement without simplification.

Carbon+Alt+Delete’s Approach to Quantifying Uncertainty

To make GHGP implementation more user-friendly, Carbon+Alt+Delete applies its principles with practical simplifications:

• Uncertainty is entered at the data point level.
• Quantified uncertainty is output at the activity level.
• Pre-defined ranges are applied, reducing the need for custom distributions.
• The system avoids overfitting and ‘uncertainty of uncertainty’.

Mathematical Foundation of the Model

The model relies on a few core statistical concepts:

• A lognormal distribution is assumed for all parameters.
• Geometric Standard Deviation (σGSD) represents the spread of possible values.
• Parameter-level uncertainty is propagated across the inventory using sensitivity analysis.

Real-World Implementation and Confidence Ranges

To make uncertainty data digestible, Carbon+Alt+Delete uses pre-set confidence intervals. For example:

• Very Good: -5% to +5%
• Good: -20% to +25%
• Fair: -35% to +60%
• Poor: -55% to +130%

These ranges allow you to generate a meaningful and conservative estimate of uncertainty without complex calculations.

Summary

Quantifying uncertainty in carbon accounting is not just a theoretical exercise—it’s essential for clear communication, robust reporting, and responsible action. By applying a simplified but rigorous version of the GHGP’s quantitative methods, Carbon+Alt+Delete enables practitioners to turn ambiguity into insight.

Even when starting with only qualitative data, the tool can generate quantitative uncertainty estimates that are defendable and useful for decision-making.