Ammonia Monitoring in Northern Ireland
UKPIR04

June 2004

EXECUTIVE SUMMARY

  1. Atmospheric NH3 concentrations were monitored at 11 sites across Northern Ireland for one year (Apr 03 to Mar 04), and at a 12th site for 7 months (Sep 03 to Mar 04). Monitoring data (both gaseous NH3 and aerosol NH4+) from 3 existing UK National Ammonia Monitoring Network (NAMN: www.edinburgh.ceh.ac.uk/cara) sites were also used to complement the measurement data from this project.
     
  2. Several of the monitoring sites were deliberately located on the edge of Areas of Special Scientific Interest (ASSI), to assess potential impacts of NH3 emissions and deposition. Other sites were located in mixed agricultural landscapes to identify representative NH3 concentrations. Two other sites were located within 500 m of two existing UK NAMN monitoring sites, which aimed to give an indication of local variability in NH3 concentration in the 5 km grid squares containing the sites.
     
  3. Air sampling was carried out using CEH ALPHA samplers and monitoring was with a monthly frequency, which provided sufficient resolution to examine the main seasonal trends and to determine the annual mean concentrations at each site.
     
  4. Gaseous NH3 showed a spatially variable field across Northern Ireland, whereas NH4+ (from the 3 NAMN sites) showed a spatially even concentration field, in broad agreement with the FRAME model, although the measured aerosol concentrations were somewhat larger than estimated by FRAME.
     
  5. Ammonia concentrations from the different sites were temporally distinct, related to the emission sources in the vicinity. Ammonium aerosol concentrations on the other hand, showed very similar temporal trends.
     
  6. There was good agreement between monitored NH3 concentrations (annual mean) and FRAME estimates for Northern Ireland, but monitored concentrations were on average lower than estimated by FRAME. A distinction may be made between the sampling points located in woodland or semi-natural vegetation (most of these sites were ASSIs) and those located in the middle of mixed agricultural landscapes. The monitoring sites in woodland or semi-natural vegetation gave lower NH3 concentrations in comparison to FRAME than the sites located in mixed agricultural landscapes. This is consistent with the sites in mixed landscapes being more representative of the 5 km grid square as a whole, as estimated by FRAME.
     
  7. It is noted that, in the past, the measurements of NH3 at the three NAMN sites in Northern Ireland (1997-2003) were around half of earlier FRAME estimates for 1996. In considering these three sites with the present measurements and the more recent model estimates from FRAME, it is noted that there is still significant discrepancy, but the deviation between the model and measurements is smaller than the previous comparison. The largest change is due to improvement in the model estimates (related to improved emissions and improved model treatment), with the measurements at these sites in April 2003-March 2004 being within ±15% of the long term mean (1997-2003). By increasing the spatial coverage of Northern Ireland (12 sites instead of 3) in this study, any bias in monitored concentrations due to nearness to source were averaged out, and this is the reason for the overall good agreement between model and measurements, despite the observed discrepancy for the three NAMN sites.
     
  8. Comparison was also made with estimates of NH3 concentration according to the NEGTAP approach, which includes a calibration of NH3 concentrations estimated by FRAME with data from the UK-wide Ammonia Monitoring Network. Comparison of these calibrated estimates with the measurements here, showed good agreement for the sites in woodland or semi-natural vegetation, and an under estimation by NEGTAP of sites in mixed agricultural landscapes. This suggests that the NEGTAP estimates are reliable for use in comparisons with critical loads for woodland or semi-natural vegetation, but may underestimate the overall budget of dry deposition for Northern Ireland.
     
  9. The measured concentrations of NH3 at the sites were combined with estimated deposition velocities for woodland and low semi-natural vegetation from the NEGTAP approach to estimate annual NH3 dry deposition. These estimates were then combined with NEGTAP estimates of wet deposition (NOy and NHx) and dry NOy deposition. Estimated N depositions at the 12 monitoring sites across Northern Ireland were 11.5 (Lough Navar) - 96 (Pantridges Farm) kg N ha-1 yr-1 for deposition to woodland (where present) and 7.0 (Lough Navar) - 56 (Pantridges Farm) kg N ha-1 yr-1 for deposition to low semi-natural vegetation (where present).
     
  10. For monitoring adjacent to the ASSI sites in Northern Ireland the range was from 26 (Deroran Bog) - 73 (Orritor, Upper Ballindery River) kg N ha-1 yr-1 for deposition to woodland and from 16 (Deroran Bog) - 42 (Orritor, Upper Ballindery River) kg N ha-1 yr-1 for deposition to low semi-natural vegetation.
     
  11. A comparison of these site estimates was made with recent updates using the NEGTAP approach (for 1999-2001) as estimated by CEH. In the NEGTAP approach, the FRAME estimates of NH3 concentrations calibrated to the overall results of the UK monitoring network are used to estimate dry deposition, with wet deposition and NOy deposition derived from a combination of inferential modeling based on other network measurements. The difference between the NEGTAP estimates and those estimated here is therefore that the present estimates use real site-based NH3 data. Based on the comparison of measured NH3 concentrations with NEGTAP estimates (bullet 8, above), it was found that the NEGTAP and site estimates of N deposition agreed well for sites in woodland or semi-natural vegetation, but that the NEGTAP values underestimated deposition for sites in mixed agricultural landscapes.
     
  12. Critical loads for nitrogen deposition have been estimated using a wide range of approaches, but it is most simple to compare the present results with the "empirical critical loads" as revised by the UNECE in 2003. In this latest revision, the critical loads are set at:
  13. Although an assessment of the exact critical loads relevant for each of the sites is beyond the scope of this study, the above range of values indicate that the nitrogen critical load is substantially exceeded at most of the monitoring locations of this study. Taking an indicative critical load value of 15 kg N ha-1 yr-1 would suggest that the only site where the critical load is not exceeded is Lough Navar.
     
  14. A simple uncertainty assessment based on the NH3 dry deposition input was made assuming a ±30% uncertainty in the deposition velocity (the uncertainty in the wet and NOy deposition of NEGTAP was not assessed here). Based on these uncertainty estimates the lower estimate of N deposition was larger than the indicative 15 kg N ha-1 y-1 critical load for all the sites except Lough Navar (deposition to low semi-natural vegetation: 7.0 [6.3, 7.7] kg N ha-1 yr-1) and Deroran Bog (deposition to low semi-natural vegetation: 15.8 [13.2, 18.4] kg N ha-1 yr-1). However, if the critical load for bogs is applied for Deroran Bog, the critical load is also exceeded at this site.
     
  15. The results show that non-NH3 deposition in Northern Ireland contributes a significant amount to the N deposition at these sites: 7 (Lough Navar) - 14 (Castle Enigan) kg N ha-1 yr-1 for forest and 4.7 (Lough Navar) - 10.8 (Castle Enigan) kg N ha-1 yr-1 for low semi-natural vegetation. However, these values may be considered rather low by comparison with other parts of the UK, while the dry NH3 inputs are rather large. This is due to the significant intensive agricultural activity in Northern Ireland, contrasting against a rather low emission of other pollutants.
     
  16. For comparison, the NH3 dry deposition in the sites in Northern Ireland ranged from: 4.5 (Lough Navar) - 82 (Pantridges Farm) kg N ha-1 yr-1 for woodland and 2.4 (Lough Navar) - 46 (Pantridges Farm) kg N ha-1 yr-1 for low semi-natural vegetation. Dry NH3 deposition contributed between 39-85% of total N deposition to woodland (mean 70%), and between 34-82% of total N deposition to low semi-natural vegetation (mean 62%). This shows that, on average, NH3 dry deposition is the species that dominates N deposition in Northern Ireland.
     
  17. The magnitude of N deposition, the dominating contribution of ammonia and the comparison with critical loads at these study sites combine to suggest that sensitive habitats in Northern Ireland are under significant threat from ammonia emissions. The present assessment has included both sites in mixed landscapes and measurements on actual ASSIs. Significant adverse effects are therefore expected on the integrity of the designated ASSIs considered in this project. The nature of the effects expected include changes in plant species composition, such as increases in grasses and a decline in heather species, sensitive woodland herbs, mosses and lichens.

Copies of this report are available from the Foundation, in electronic format on CDRom at £20.00 + VAT or hard copy at £15.00, less 20% to FWR members.

N.B. The report is available for download from the SNIFFER Website