The high level of uncertainty associated with the PAH inventory is largely due to the lack of actual measurements from potentially significant sources. To significantly improve the quality of the inventory requires the measurement of emission factors and detailed assessment of the processes involved in the following areas:

• road vehicles, both diesel and petrol driven
• off road fuel use
• coke production;
• creosote use;
• open burning;

• other domestic and industrial combustion

As with PAHs, the uncertainties associated with the emission of PCDD/F can be attributed to a lack of measurements from significant source sectors. Therefore the priority areas to further improve the PCDD/F emission inventory are :

• non-ferrous metal production;
• chemical industry
• accidental fires
• wood combustion
• iron and steel

The emission inventory for PCBs is very uncertain as the information on the quantity of PCBs in electrical equipment is based on an estimate. In addition, leakage rate from these dominating sources is also estimated.

Emissions of pesticides have been calculated by using available data from Europe or the USA. Due to the scare nature of the data, the confidence in the emission estimates is low. The emission factors used here have been derived for specific methods of application and during certain atmospheric conditions. These conditions may not represent those in the UK, but until further more applicable data become available it is difficult to reduce the uncertainties associated with these emission estimates.

Due to the lack of available data, it has only been possible to give single estimates for the emissions of SCCPs. Indeed different estimation techniques give estimates of 1.5 and 62 kg/year indicating the high levels of uncertainty.

Emissions of heavy metals are particularly uncertain owing to the variation in metal content of fuels and raw materials used in industrial processes. Most of these emissions are calculated using emission factors expressed in terms of the mass of metal emitted per unit mass of fuel burnt. For industrial processes the emission factor is in terms of mass of metal emitted per unit mass of product. These emission factors are typically assumed to be constant with time.

It is difficult to estimate the accuracy of the heavy metal emissions estimates because they are based on a limited set of measurements some of which are derived from foreign sources which may not be applicable to UK conditions. It is thought that the estimates of lead, cadmium and mercury are the most accurate, and the lowest in uncertainty. Emissions of zinc, nickel, copper, chromium and arsenic are considered to be less accurate, whilst emissions of selenium and vanadium are thought to be the least accurate, with the highest uncertainties. Wenborn et al (1998) estimated the uncertainties in heavy metal emissions in 1990 using a Monte Carlo analysis. Based on the minimum-maximum spread the uncertainties were: lead ± 14%; cadmium ± 38% and mercury ± 28%. The uncertainty in the lead emission for more recent years will be greater than this due to the lower contribution from petrol combustion.