12. Projections

Last updated on 10 Dec 2013 11:53 (cf. Authors)


For many years, the focus for emission inventory work in Germany has been on the past. The main goal was to provide complete time series of emission data for all source categories. This report deals with these efforts in great detail.
While decent data on historic emissions are key to the political process and to decisions on abatement technology promotion, future emission paths do have the power to shed a new light on these discussions. The main idea behind emission projections is not so much about exact numbers, but about the comparison of scenarios. Scenarios are made up by a distinct number of actions and measures, translated into future values for activity rates and emission factors. The questions answered by emission projection is therefore not "How many kilo-tonnes of sulfur oxide will be emitted in 2020?" but "How does scenario X/provision Y compare to the reference emission path? Does it significantly help our goal to further reduce emissions?"
The German projections have seen a major overhaul in 2010. A new emission projection database (called "emission scenarios") has been created. Multiple scenarios are taken into account, sketching development of activity data and emission factors up to 2050. The new system features integrated assessment for both greenhouse gases (GHG) and air pollutants. In particular, existing projections for GHG can be applied to air pollution contexts. The databases also allows for the flexible combination of distinct scenarios for specific source categories to add up to a complete inventory. Today, the projection database is fully operational and used as the common basis for reporting on emission projections.


The main German emission inventory uses a database called "Central Emission System" (CES). The CES holds very detailed information on activity data (AD) and emission factors (EF) for GHG, air pollutants, heavy metals, and persistent organic pollutants. To give some number: there are about 2,000 AD, 20,000 EF, and 20,000 emission time series' in the CES. All these data together represents the German inventory for historic emissions.
For projection purposes, this structure is too detailed. Projecting future trajectories for such a high number of time series has proven to be very difficult, if not impossible. Therefore, the new projection system applies some aggregation techniques. The number of AD time series' is boiled down to 153, for example. This is mainly achieved by discarding source category distinction: for the most important sectors "Energy" and "Industrial Processes" the system splits data into 13 instead of hundreds of sub-categories. The aggregation is automatically fed by the CES and can be easily updated upon changes in the main database.
By default, all data from the CES for historic years is collected and put into the projection database. In the "Reference" scenario (REF, see below) most time series are set to extrapolate the last given value, i.e. all future emissions will remain at the same level as the last historic value. Other scenarios are free to change any of the extrapolated values of the REF to reflect their "storyline". The figure below illustrates this workflow.


The current version of the projection system includes five scenarios. While details are explained below, this section looks on the common meta data used.


Every projection depends on its parameter inputs. Scenarios are little more than umbrellas for values put in place for parameters taken into account when modelling future development. All results reflect the input settings and are only "valid" if the projected inputs match the actual development. On the other hand, no scenario is "in-valid" per se - it may be improbable, but still interesting to study. In particular, when looking for sensitivities, scenarios outside common predictions serve well.

The following table shows some of the most important parameters taken into account for emission projections.

Parameter Description
Demographics Future population and household numbers obviously effect emission levels to come.
Economics The 2009 economic crisis clearly demonstrated the link between economic growth or disruption and the amount of pollutant released.
Fuel prices For most pollutants fuel combustion is a mayor contribution to emission levels. Since the amount and composition of fuels used depends on their price, emissions in turn depend on fuel prices.
Legislation For many pollutants specific regulation has been introduced, some of which is expected to show effects for future emission factors, e.g. vehicle or power plant regulation.

Each of the scenarios used specifies its own set of values for these parameters.

Economic crisis 2008/2009

In 2009, all projection work done prior its start in late 2008 seemed invalidated due to the economic crisis. Real emissions of 2009 have been way below projected values. But, at least in the case of Germany, the re-acceleration of economic dynamics made 2010 numbers raise again. Most are now a little below 2008 values and therefore just inside most projections uncertainty ranges. Unless the next part of the crisis is still ahead of us, most projections can still be considered applicable.


The emission projection system has been recently updated and now covers four scenarios:

  1. Reference scenario (REF): The REF scenario represents an upper bound projection. For most source categories this would usually mean that emission do not decline, but remain stable on the level of the last reported year. Both activity data and emission factors are extrapolated as constant values into the future. There are some exceptions, most notably aviation, which are excepted to show raising emission levels in the future. Consequently, their activity values do also climb in reference scenario. From these semantics it can be inferred, that the REF emission projections should always be higher then all (or at least most, if you consider an "economy boost" scenario) other scenario levels. This does not work perfectly for crisis years like 2009, because the REF scenario does not anticipate the rise in emissions following in 2010 (see graphs below).
  2. Emission working group scenario (I26): The I26 scenario represents a "most probable" projection. The emission inventory working group at the Umweltbundesamt (called I2.6) frequently updates these data as new information comes available. Therefore, this scenario is usually the first to include rapid or unexpected developments such as the economic crisis in 2009. For the upcoming year data from this scenario can is considered to be very close to the actual values.
  3. Current policy scenario (PSzVI-APS, [1]): The current policy scenario and the structural change ("Energiewende") scenario - see below - are adopted from the most current version of the energy modelling and projection project "policy scenarios (Politikszenarien)", carried out in the context of climate change reporting. The results of this study include detailed future activity data, fit for purpose of the emission scenario tool. The current policy scenario takes into account all measures that were in place by July 2011.
  4. "Energiewende" scenario (PSzVI-EWS, [1]): The structural change scenario includes all the measures of the APS and adds some more on top which are currently in discussion and/or preparation.


The images below represent a small selection of the assessments produced by the projection system. These results are also a regular part of the German data submission under the UNECE/CLRTAP, please refer to the corresponding tables for details.

Next section: 13.1 Point Sources

1. Umweltbundesamt, 2013: Politikszenarien für den Klimaschutz VI (to be released shortly)
Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-NonCommercial-NoDerivs 3.0 License