Which kinds of uncertainty does the Greenhouse Gas Protocol Corporate Standard say a company can measure, and is a change between two years more certain than the total?

Question

Accepted Answer

The Greenhouse Gas (GHG) Protocol Corporate Standard splits uncertainty into scientific and estimation kinds, and says only parameter uncertainty is within reach for most companies. If a plant uses the same method each year, its systematic errors are for the most part identical, so the change between years is generally less uncertain than the total. Scientific uncertainty: the science of the emission or removal process is not fully understood. The guide's example is the factors behind global warming potential (GWP) values. It says quantifying this is likely beyond most company inventory programmes. Estimation uncertainty arises any time emissions are quantified, so every estimate has it. It has two parts. Model uncertainty: the equations used are wrong, or the inputs to them are inappropriate. Estimating it is likely beyond most companies, though some may use their own engineering expertise. Parameter uncertainty: how well the inputs, meaning activity data and emission factors, are known. It can be judged by statistics, instrument precision and expert judgement. The guide says it will be the primary focus of companies that investigate uncertainty. For most inputs, such as litres of petrol bought or tonnes of limestone used, only one data point exists. Where statistics are missing, companies will usually have to rely on expert judgement, which the guide says is difficult to obtain in an unbiased and consistent way. So almost every full uncertainty estimate is imperfect, has a subjective part and is itself highly uncertain. The guide concludes that in most cases uncertainty estimates cannot be read as an objective measure of quality, and cannot be used to compare the quality of estimates between source categories or between companies. The guide names two cases where comparison works. In the first, two similar facilities use identical methods, so differences in scientific and model uncertainty can for the most part be ignored. In the second, a single facility uses the same method every year. Then the systematic parameter uncertainties, as well as the scientific and model ones, are for the most part the same in both years and cancel out. The uncertainty in the difference between two years is therefore generally less than the uncertainty in one year's total. Such trend estimates can track relative changes in the quality of a facility's estimate and can guide a facility's reduction target. The guide says they are likely to be less useful for broader, company-wide targets, because of the problems of comparing uncertainty across gases, sources and facilities. Keep the method and factor set the same each year and write them down, so your yearly change rests on the firmer case. If you must change the factor source or the method, recalculate the earlier year on the new basis so the comparison stays like for like. Our reading: a change in factor source breaks the cancelling, so it can turn a real cut into an apparent one.

Which kinds of uncertainty does the Greenhouse Gas Protocol Corporate Standard say a company can measure, and is a change between two years more certain than the total?

The kinds of uncertainty

What the numbers cannot do

Why a yearly change is firmer than a total

What to do with this

A factor that is 5 percent too high in both years (example figures)

What Senseible does here