This element forms a major part of the contract through assisting DETR to develop the UK's negotiating stance within the context of both the UNECE and the EC. A range of scenarios have been run using EMEP area emissions data from IIASA interim reports 5 to 7 and papers prepared for the Working Group on Strategies. These scenarios are summarised in Table 5. The spatial distribution of UK emissions in 2010 has been based on two different estimates, an old estimate applied from the start of the contract until the summer of 1999 and a new estimate developed for work towards the EU's Emissions Ceilings Directive (see below).

Between April and June 1997, HARM10.4 was used to provide scenarios for the development of the EU Acidification Strategy. These related to the B1 scenario (50% ecosystem area gap closure). Two variants of B1 were run: B1 flat and B1 fgd. The former assumed that there was no coal fired electricity generation in the UK, while the latter had four coal fired stations operating. Output from these model runs was supplied to ITE Monks Wood for CL mapping. Between July and September the new 10 km version of HARM was used for a series of model runs stepping down UK only and UK and EMEP area emissions between REF and B1 emissions. Model output from current emissions, REF and B1 were supplied to ITE Monks Wood. Scenario data were also supplied to Chris Curtis and Tim Allott (UCL) to run within the FAB model. Over the same period, the 20 km model was used to explore the effects of changing the S content of liquid fuels (for Sarah Bolt).

HARM has been used in the past to model concentrations of secondary aerosols (QUARG, 1996). Between April and June 1998, the 10 km version of HARM was used to model these secondary particulates as part of work for APEG. Data were supplied to Helen ApSimon (Imperial College) and Dick Derwent (Met. Office).

As new reports are made available by IIASA, so the emissions used in HARM and ELMO scenarios can be updated. HARM11.3 was used to produce a number of scenarios (1990, REF and E10/1) based on IIASA 5^th Interim report. The same scenarios were also run but based on cost curves derived from ASAM rather than RAINS. It should be noted that these scenarios were affected by an error in the EMEP area NH₃ emission factor which was corrected in HARM11.4. HARM11.3 was also used to model SO₂ concentrations in 2005. These data were supplied to NETCEN for work on the UK's Air Quality Strategy. Using HARM11.4, new scenarios were run based on IIASA's 6^th Interim report (October to December 1998). These were: a new REF scenario; F8 and MFR. Due to HARM's underestimation of dry NH-N deposition, the dry reduced N deposition fields were supplied by FRAME output for the same scenarios. Further HARM11.4 scenarios were run early in 1999: post-Kyoto; high ambition; H1, J1 and G5/2. The same model was used to model a 2005 scenario taking in to account possible new Large Combustion Plant standards. Both current commitments and the application of LCPD to existing plant were considered. Emissions data for these runs were supplied by Colin Powlesland of the Environment Agency.

Over the summer of 1999, HARM11.5 was used to run a wide range of scenarios for the development of the EU Emissions Ceilings Directive. In these scenarios HARM output was used exclusively (no FRAME results) following the implementation of the alpha factor approach in modelling dry NH-N deposition (see Table 4). Scenarios H1 and J1 were run using both a previous disaggregation of UK emissions and a new distribution which took into account both the terms of the Kyoto Protocol and the Coal Review. A Baseline scenario (now called UK-REF) and UK1 (later called the 4^th scenario) were run using the new emission distribution, the run for 1990 was also repeated. A series of scenarios (UK7, UK14 and UK15) were run which kept all emissions as J1 except for UK NH₃. Modelled depositions were supplied to ITE Monks Wood and gas concentration and aerosol data to Imperial College and AEA Technology. Details of these and the associated ELMO scenarios were presented to the Department in a report 'Costs and benefits for the UK of complying with the EC National Emissions Ceilings and Ozone Directive and the UNECE multi-pollutant, multi-effect protocol' coordinated by Mike Holland of AEA Technology.

HARM was used to model the WGS31b scenario (IIASA report September 1999). The UK's emissions of SO₂, NOx and NH₃ were the same as those agreed for the Gothenburg protocol signed in December 1999.

Output from HARM11.5 for 1990, 1995, the baseline scenario (UK-REF) and a representation of the Gothenburg protocol is shown in Figure 5 for S and total N (oxidised plus reduced). These model runs all used a 1961-90 rainfall field. The figures highlight a substantial reduction in S deposition, but a smaller reduction in N deposition.

At the request of Claire Skipsey, HARM11.5 was used to estimate the effects of changes in the S content of liquid fuels. The output was supplied to Alastair Ritchie of NTEC.

HARM is being used extensively within NEGTAP. An initial series of model runs was carried out to explore changes between SO₂ emissions and S deposition. So far this has been done for 1970, 1980, 1983 and 1995. In this case, year specific rainfall field are used. Further runs will be carried out when the emissions and precipitation data are available.

Early scenario work using ELMO focused on assessing the response of peak ozone concentrations to changes in CO, VOC, isoprene and NOx emissions across the UK and the EMEP area. In the period April to June 1998, ELMO was used to model O₃ concentrations resulting from 50% flat reductions in UK and EMEP NOx and VOC emissions. The results were supplied to John Stedman at NETCEN. Other ELMO scenarios included: 1995 (the base year for ELMO); 2010; E10 and 2 variants of this.

As part of the work with AEA Technology on the Emissions Ceilings Directive, ELMO was run for baseline (UK-REF), J1, H1 and UK1 (the 4^th scenario). Output was modelled for peak O₃, EPAQS and WHO exceedances, AOT40, AOT60 and secondary aerosols. Data were supplied to AEA Technology and Imperial College. Modelled peak ozone for 1995, baseline (UK-REF) and a representation of the Gothenburg protocol commitment is illustrated in Figure 6. It should be noted that as a result of a change in the way in which the UK VOC emission inventory was presented, emissions of isoprene were double counted in these runs. The model has been re-run with the correct value, but the change made little difference to the modelled ozone concentration.

One issue which has been raised by recent scenario work is that of the difference between the declared UK total emission and mapped emissions. Mapped emissions (as supplied by the NAEI) exclude some source categories (e.g. off shore installations) and so are slightly lower than the listed totals. It would be preferable for mapped and total emissions to be the same. There has also been a change in the way in which EMEP present the emissions of the wider area. The domain of the EMEP model has been enlarged and country codes are not directly assigned to the emissions data files available over the www. Emissions from a few areas do not have country codes at all. In the past, basic EMEP area emissions data have been supplied by EMEP with country codes which has allowed their easy integration in to HARM. It now seems that data from 1997 on will have to be processed before they can be used in HARM (or other models requiring country information).