Lowcost methods for gasparticle distribution of nitrogen species
Low-cost methods for gas/particle distribution of nitrogen species Mark Sutton, Sim Tang, Eiko Nemitz, Neil Cape, David Fowler Centre for Ecology and Hydrology (CEH) Edinburgh
Measurements of what? • Gas concentrations of reactive nitrogen – HNO 3, NH 3 • Aerosol concentrations of reactive nitrogen – NO 3 -, NH 4+ • Added species (in parallel, some methods) – Gaseous SO 2, HCl – Aerosol SO 42 -, Cl-, base cations…
Why speciate between gas and aerosol? • Spatial patterns of gas and aerosol differ – Gas concs vary more locally – Sites may not be representative for gases – The transboundary contribution to aerosol concs can be larger than for gases • Temporal trends of gas and aerosol differ – With reductions in S, gas-aerosol partitioning has changed. – e. g. NH 4+ aerosol concentrations have decreased substantially, while NH 3 has not decreased much.
Why not just rely on daily filter-packs? • Method well established in EMEP • But are they really low cost? – Low capital cost, but high running cost for daily sampling. • Accept the need for high temporal resolution data – to examine relations to air mass origin at selected sites • But daily filter packs are expensive way to collect data of questionable quality – Well known artefacts between gas and particle phases
Example of artefacts with filter-pack sampling of NH 4+ aerosol FP and Berner Impactor compared against Steam Jet Aerosol Collector (SJAC) Warm conditions Negative artefact Cool conditions Positive artefact Nemitz (1998) Artefact increases with sampling time
Filter-pack data only for total TIN & TIA Eskdalemuir, UK Which part is gas & which part aerosol?
High-cost solutions (Levels 2 and 3) • Daily annular denuder-filter combinations (e. g. Chemspec system) • Continuous denuder and Steam Jet Aerosol Collection methods • Strong desire for hourly AMANDA continuous denuder frequency or better • See next presentation
Low-cost solutions (Level 1) • What is low cost? – Low capital cost (<€ 1 k per site) – Low running cost (<€ 5 k per site per year) – Typically monthly sampling • What is the role of these methods? – – Seasonal and inter-annual trends Spatial patterns with high site density Dry deposition Contribute to QA/QC of other methods
Example methods • Passive samplers – Can work well for NH 3, but can also fail spectacularly – Methods must be continuously referred to an active sampling reference – Not available for HNO 3 or aerosol • Long-term denuders – Provide a cheap reference method for all main gas and aerosol species
Passive sampling for ammonia • Most passive methods can be made to work if well operated and calibrated • Methods fail when not tuned for detection limits and when not properly calibrated (e. g. results out by factor 10!) • Many methods available, Gradko, IVL (Ferm), CEH (Alpha) Replication of sampling (min 3) essential for QC ALPHA sampler Key (all samplers are cylindrical): Sampler body Sampler cap Sampler membrane Acidified filter/mesh Circular supporting inserts
Example of change in passive sampling method – Stoke Ferry Gradko tubes Alpha samplers Data using UK wide calibration for passive methods
Long-term denuders (DELTA) DEnuder for Long-Term Atmospheric sampling • Tuned for monthly sampling: Good for local site operators and allows low sampling rate (0. 3 l min-1) • Low sampling rate allows short (10 cm) denuders: Robust system can be sent by post to site operators) • Two denuders in series to allow QC on capture efficiency for every sample • Post-denuder filters to sample aerosol (first basic, second acidic)
Denuder for Long-Term Atmospheric Sampling (DELTA) • Cheap ~€ 800 for each complete system • Monthly sampling 50 times cheaper than daily monitoring • Double system (4 denuders in series) to include acid gases and other aerosol (SO 2, HNO 3, HCl, SO 42 -, NO 3 -, Cl- etc)
Architecture of full DELTA system 14 cm long denuder coated with K 2 CO 3 10 cm long denuder coated with citric acid Collect ammonia Collect acids 14 cm long denuder coated with K 2 CO 3 10 cm long denuder coated with citric acid Collect anion aerosol 25 mm paper filter in K 2 CO 3 25 mm paper filter in citric acid Denuders from 6 mm i. d. borosilicate glass Collect NH 4+ aerosol
Example HNO 3 and NO 3 - data from full DETLA system
HNO 3 network: gas & aerosol concentrations 2002 HNO 3 HCl SO 2 Na+ NO 3 - Cl- SO 42 - Ca 2+
Interpolated monitored NH 3 and NH 4+ concentrations Methods: DELTA (monthly denuder/filter pack) ALPHA (passive) at additional sites for gas only
Dry deposition monitoring • Time Averaged Gradient (TAG) system • Utilizes the DETLA approach with conditional sampling for seasonal dry deposition estimates
How does the DELTA method compare with the daily FP data?
Can we detect the return of livestock after foot & mouth disease in the NH 3 record? Study motivated by Dutch & E. European experience of difficulty to see trends following emission reductions Added Alpha sampling at ~20 targeted locations Monitoring sites and results of prior modelling
Comparison of affected & unaffected areas around Cumbria 2002 Monthly modelling misses December peak It is only possible to detect the trend due to use of multiple sites 2003
Conclusions • Low-cost denuder methods provide a robust and effective reference – including QC • Passive sampling can work for NH 3, but not for aerosol • Use for regional monitoring of long term trends • DELTA method provides a tool for QA/QC of daily FP where that is used.
For further discussion • Practical considerations • Denuder coatings for warm climates (e. g. phosphorous acid) • Variant designs • DELTA as reference method for passive approaches (e. g. CEN TC 264/WG 11) • Agreeing data quality objectives • Establishing representativity of L 2 and L 3 sites in 50 km grids.
Importance of in-square variability Dry NH 3 5 km 50 km Wet NH 4 NH 3 conc
Similarity of NH 4+ at adjacent remote sites in N. Scotland
TIN and TIA at High Muffles
SO 2 and SO 42 aerosol at Eskdalemuir
- Slides: 28