2.7 General Uncertainty Evaluation

Last updated on 24 Jan 2013 17:05 by Kevin Hausmann

As a first approach for uncertainty analysis of the German inventory the results of a research project are presented.
The PAREST project (“Particle Reduction Strategies” - UFOPLAN FKZ 206 43 200/01 - Strategien zur Verminderung der Feinstaubbelastung) is a research project, in which emission scenarios until 2020 were constructed for fine particles (PM10 und PM2.5) as well as aerosol precursors SO2, NOx, NH3 and NMVOC, both for Germany and Europe. Reduction measures were assessed and finally air quality in Germany was modeled.

For PAREST, all uncertainty input variables to the German Federal Environmental Agency’s model system of emission reporting were replaced by probability distribution functions. The parameters determining the shapes and quantiles of the functions were taken over from primary and secondary literature or estimated through expert judgement.

Based on that, a Monte-Carlo simulation was carried out to analyse the uncertainties of the German nationally aggregated 2005 emissions of fine particles (PM10 und PM2.5) and of the aerosol precursors SO2, NOx, NH3 and NMVOC.

Pollutant - [%] + [%]
PM10 16 23
PM2.5 15 19
NOx 10 23
SO2 9 9
NMVOC 10 12
NH3 13 13

Looking at the results (95%-confidence interval), the inventories for PM10 (-16%/+23%), PM2.5 (-15%/+19%) and NOx (-10%/+23%) appear most uncertain, while the inventories for SO2 (-9%/+9%), NMVOC (-10%/+12%) and NH3 (-13%/+13%) show a higher reliability.

The source categories adding the most relevant contributions to overall uncertainty vary across the pollutants and comprise agriculture (NOx from fertiliser application, NMVOC from manure management, NH3 from animal husbandry and cultivation of land, PM10 from pig fattening), mobile machinery in agriculture and forestry (PM10, PM2,5 and NOx), construction sites/2.A.7.b (PM10), small businesses /carpentries (PM10 and PM2.5), cigarette smoke and fireworks (PM2.5), NOx from heavy duty vehicles and passenger cars, NMVOC and NH3 from petrol engines), solvent use (NMVOC) and stationary combustion (SO2 from coal-fired power plants and oil-fired domestic furnaces, PM10 and NMVOC from wood firing).

As an additional information the following table shows the uncertainties in our database CSE for relevant pollutants of the source category agriculture, the first source category with a complete uncertainty data set for NEC pollutants and PM for the current inventory, compared with the sectoral uncertainties (higher value of the result of the Monte-Carlo simulation) for agriculture from the PAREST project.

Pollutant Agriculture Uncertainties PAREST project Agriculture Uncertainties CSE database
NH3 14 % 10 %
NOx 420 % 402 %
PM10 58 % 65 %
PM2.5 67 % 72 %

A quantitative uncertainty assessment for the current inventory should follow for all categories as soon as updated uncertainty data will be available. The determination of uncertainty values for activity data, emission factors and emissions is implemented in every running and planned research project, so an informative uncertainty data set for a number of categories will be expected in time. Available uncertainties of activity data from UNFCCC reporting and defaults of the "EMEP/CORINAIR Guidebook" and the “IPCC Good Practice Guidance” will be used as well.

Next section: 2.8 General Assessment of Completeness

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License