Mass Closure

The status of the data reduction related to mass closure on the ship and at Cape Grim was reported by Quinn and Zhuang, respectively.

Ship mass closure. Several independent measures of mass have been compared. These include:

• ionic mass which is the sum of the ion mass concentrations as determined by ion chromatography for the sub- and supermicron fractions
• mass derived from gravimetric analysis of impactor stages for the sub- and supermicron fractions where the gravimetric analysis was done in an NH3-free, humidity-controlled glove box at 33% RH
• mass derived from the number size distribution where a) the number distribution was converted to a volume size distribution and b) density derived from the 7-stage impactor measurements was used to convert from a volume to a mass size distribution. A chemical equilibrium model was used to determine the density of each impactor stage.

Results:

At this point, the mass closure calculations have been done for a low reference RH of 30%. On average, the ionic mass accounted for 85% of the submicron gravimetrically-determined mass. A few measurements of the total organic carbon content in particles less than 300 nm in diameter indicate very low concentrations (3 to 10 ng/m3) such that it does not appear to contribute to a significant fraction of the submicron mass. The supermicron ionic and gravimetrically-determined mass agreed within experimental uncertainty.

In addition, in 16 out of 20 cases, the gravimetrically-determined and number-derived mass agreed within experimental uncertainty. The 4 cases of disagreement correspond to time periods when continental air was sampled. Agreement between the gravimetrically-determined and number-derived supermicron mass agreed within experimental uncertainty for all cases.

One further comparison of gravimetrically-determined mass was made wherein impactor samples collected at the base of the sampling mast at 50% RH were compared to open face filter samples collected outside near the inlet of the sampling mast. This comparison was made to determine how much mass is lost in the heated inlet. In 8 out of 12 cases, the submicron mass at the bottom and top of the sampling mast agreed within experimental uncertainty. Disagreement in the submicron mass could be a result of an uncertain size cut for the open face filters (8 um pore size Nuclepore filter followed by a 1 um Miliipore filter) at the top of the mast. Only 3 out of 12 cases showed agreement between the supermicron mass collected at the bottom and top of the mast. In all of these cases of disagreement, the mass collected at the top of the mast was larger than that collected at the bottom. In addition, the disagreement was larger during the latter half of the experiment.

Future work:

Future work will involve performing the mass closure calculations at a higher reference RH more relevant to the atmosphere. This is expected to improve the agreement between the submicron gravimetrically-determined and ionic mass. In addition, for those cases where disagreement was largest between the submicron gravimetric and ionic mass, the samples collected for total organic content will be analyzed. An attempt may also be made to quantitatively define the loss of supermicron mass within the sampling inlet by determining particle losses associated with the open face filter as a function of wind speed and direction across the filter (Quinn/Coffman).

Cape Grim Mass Closure

To date, two independent measures of mass have been compared. These include:

• ionic mass derived from multiple-stage MOUDI samples
• mass derived from gravimetric analysis of multiple-stage MOUDI samples where the gravimetric analysis was done in a humidity-controlled glove box at 10 to 20 % RH. aluminum substrates were used for the sample collection.

Results:

This comparison indicates that a large (up to a factor of 5) fraction of the gravimetrically-determined mass is unaccounted for by the ionic analysis across the entire size spectrum. The PALMS data indicate that organics were present in the Cape Grim aerosol but these data do not indicate the particle size or concentration of the organic matter. Samples collected for total organic and elemental carbon were analyzed and showed extremely low values. These samples were stored for a year before analysis, however, such that much of the volatile organic matter could have been lost. (It is not clear at this point if the samples were stored frozen).

Experiment-average values of the scattering coefficients measured on the ship and at Cape Grim for both the submicron and total aerosol (Dp < 10 um) show remarkable agreement. This agreement indicates that the aerosol mass concentration was similar at the two platforms and that, perhaps, the gravimetrically-determined mass at Cape Grim is too high.

Future Work:

During the working group discussion, many suggestions were made to improve the Cape Grim mass closure. These include:

Other independent measures of mass

1. Estimate mass from DMPS, APS, ASASP collected from CAI (Howell/Gras)
2. Complete gravimetric analysis on submicron filters collected during ACE 1 from the CAI by CSIRO (Gras)
3. Calculate mass scattering efficiencies (light scattering coefficient/unit mass of ions or gravimetrically-determined mass). Using density estimated from chemical composition, compare resulting scattering to measured scattering (Rood)

Other methods for determining unaccounted-for mass

1. Use Al foils that were sampled during ACE 1 for gravimetric and/or speciated organic analysis by U Hawaii and extract, filter, and analyze for the insoluble fraction(Howell/Anderson)
2. TSP data from atomic analysis exist for 4 to 5 years in the early 1980s. Is there evidence in this data record for large concentrations of insoluble material at Cape Grim? (Gras)
3. Analyze CSIRO filters collected for gravimetric analysis chemically (Howell/Gras)
4. Use Cainey's MOUDI data for organic acids to determine what fraction of the mass they contribute
5. Use PM 2.5 samples analyzed by PIXE for an indicator of soil dust (Gras)

Re-check of gravmetrically-determined or IC mass

1. Investigate possibility of Aluminum oxides forming on impactor substrates during sample collected for gravimetric analysis (Zhuang)
2. IC analysis of MOUDI foils--did they pick up NH3 (g) (Zhuang)