130 likes | 144 Views
Explore the limitations and uncertainties in PM measurement and modelling, including the underestimation of concentrations and large uncertainties in emissions inventories. Discover the need for improved qualification of measurements and the challenges of linking sources to observed PM levels.
E N D
Working group II Questions related to PM measurement and modelling
General remarks • All models underestimate PM concentrations; no mass closure • Large uncertainties in emissions inventories • Need for a better qualification of what is expected from measurements : • too few observations for too many parameters • technical difficulties • monitoring strategy still need to beelaborated in the networks
Q7 : How well can we link sources to observed PM levels with atmospheric models • Secondary Inorganic Aerosols (SIA); Sulfate • Emission inventories are rather satisfactory excepted (perhaps) for ships and harbours activities • Measurements are reliable : long experience, high resolution possible • Models give relevant results for atmospheric concentrations, although improvements related to the cloud chemistry are still possible (and encouraged).Deposition is more difficult, especially wet deposition.
Q7 : How well can we link sources to observed PM levels with atmospheric models • Secondary Inorganic Aerosols (SIA); Nitrate • Few measurements in Europe, made with filter-based techniques: problem of evaporation still not solved • Good temporal resolution needed considering the processes • How to translate what is on the filter in what is modelled? Numerous and quick reactions involved. • Need for concomitant ammonia, nitric acid, NO2, nitrate measurements for modelling (intensive campaigns) • Sensitivity to the meteorological parameters: temperature, humidity • Issue for ammonia emissions, still inaccurate for agriculture
Q7 : How well can we link sources to observed PM levels with atmospheric models Acceptable results for SIA • Measurements : OK • Models : “trustable” • Improvements still possible but not considered as a priority
Q7 : How well can we link sources to observed PM levels with atmospheric models EC/OC • No reference method, lack of accuracy and lack of consistency between European measurements. German standard exists but work still needed • Large differences in results issued from different methods • Total carbon could be well estimated but modellers need to know the primary/secondary parts from anthropogenic/biogenic sources • Recommendation : Need for a standard method!
Q7 : How well can we link sources to observed PM levels with atmospheric models EC/OC • Emission of primary carbon should be improved : quantities, speciation - wood combustion, residential heating) • Improvements needed to correctly grid the emissions and to get the right temporal profiles • Recommendation : develop levoglucosan measurements (source/receptor approach) and find other tracers • VOC emission and speciation (anthropogenic and biogenic) should be improved for modelling the secondary part
Q7 : How well can we link sources to observed PM levels with atmospheric models EC/OC • Obvious and well-known difficulties for modelling: inconsistencies with measurements (but which measurements?) • Complexity of the processes • Developments still needed but they are hampered by the lack of measurement data and emission uncertainties Finally models are not ready • BUT EC/OC could represent 30 to 50 % of the PM composition!Recommendation : Priority on EC/OC measurement and emission inventories
Q7 : How well can we link sources to observed PM levels with atmospheric models Dust/ natural compounds • Measurement of representative compounds : calcium, sodium… : OK • Distinction resuspension and “real” natural sources • Resuspension : soil erosion, traffic • Parametrizations used in the models: uncertainties but global improvement of the results • Sensitivity to weather conditions, accuracy of the land cover needed
Q8 : How large are the uncertainties in PM measurements and model? • SIA : • 15-20% for sulfate modelling (annual average) • 40% for nitrate modelling (annual average). Same order of magnitude as the measurement uncertainty • Acceptable • Dust/natural compounds : • difficult to assess • non determinant • EC/OC • (Very) Large uncertainties in measurement, emission inventories and modelling Current models should not be used for IAM and policy purposes related to PM
BUT : relevant experiences with PM modelling • Recommendations • Define a “protocole” to check the reliability of the models to represent the right chemical regimes • Consider the models results as absolute values but comparing scenarios : “relative responses” • Need to concentrate measurements at relevant sites: measure all the PM components at the same site, and also gaseous compounds to reach a better understanding of the processes and the chemical regimes
Q8 : What improvements are required in PM monitoring, modelling and basic scientific understanding for the assessment of health and climate impacts of PM? • PM measurement : unbearable situation because of the heterogeneity of the measurements in Europe • Recommendation : Use methods which have shown equivalence with the reference method, on a daily basis • PM modelling : Improvement expected for meteorological data and boundary conditions • Better investigation of the urban scale modelling needed (where exposure occurs) • Recommendation : to assess and compare methods available to deal with the urban scale from the regional scale
Q8 : What improvements are required in PM monitoring, modelling and basic scientific understanding for the assessment of health and climate impacts of PM? • Promote data assimilation but be aware of the quality of the data assimilated • Promote measurement of vertical profiles (chemical and meteorological) • Recommendation : to measure ad evaluate particulate number and size distribution (for health and climate impact)