Interactive comment on “ Carbonaceous components , levoglucosan and inorganic ions in tropical aerosols from Tanzania , East Africa : implication for biomass burning contribution to organic aerosols ” by S

This paper presents database for PM, carbonaceous components, levoglucosan and water soluble inorganic ions and illustrates that biomass burning and atmospheric photochemical processes are the main sources in Tanyania. No matter the experiment or the long paper take authors’ much effort to finish the work, while there are no new scientific findings that could arouse readers’ much interest on this paper. So I am afraid it is not probable to publish on ACP. Most of the results are the same or small difference with the referenced literatures as the author mentioned in the paper. Even though some new points, such as higher nss-K+/EC ratios in Tanzania than other nations, the

Reply: The sentence is reworded in the revised MS lines 77-79 as ". ... . .levoglucosan, which is produced by the pyrolysis of cellulose, has been used as a unique molecular tracer of biomass burning aerosols (Simoneit et al., 1999;Fraser and Lakshmanan, 2000)."P4.To show the sampling site in a separate figure with parameters such as long/latitude, altitude and cities/mountains/oceans around the site would be much clear for readers to have a general impression on the local situation.In the text, popularity C12868 and industry in the city should also be explained roughly.
Reply: A Figure showing the sampling site and description of the site will be added in revised MS as follows: "Intensive aerosol sampling campaigns were carried out at a rural background site (300,000 inhabitants) in Morogoro (06 • 47'40.8"S;37 • 37'44.5"E,altitude 504 m, a.s.l) during the wet season (30 May to 13 June 2011) and the dry season (28 July to 8 August 2011).This site is located at about 200 km west of the Indian Ocean and the city of Dar es Salaam, a business capital in Tanzania (Fig. 1).The site has no major industries and the possible main local aerosol sources include field burning of crop residue and waste, emissions from livestock (cattle diary or farm) and domestic/forest fires".P5.Line 8: Whether the samples will be firstly packed in an aluminum file?Otherwise the stuff on the filter will probably stick to the upper one.
Reply: We thank the reviewer for the technical note.We would like to clarify that our sampled filters were not parked in an aluminum foil; instead after sampling they were folded in half face to face and placed in the jar during storage.This point is added in lines 119-121 in the revised MS.
P6. Line 25: A brief description of QA/QC in the method of analysing anhydrosugars should be introduced, including recovery, standards, linearity and so on.There is also some discussion of mannosan in the text, thus title of 2.2.3 section is not complete and it should be changed.
Reply: Thanks for the helpful comments.Sentences describing quality assurance and control in levoglucosan analysis were added in lines 171-176 in the revised MS.We also deleted "mannosan" in our discussion in the revised text.P10.Line 1: Could you please explain the reason why your PM10 mass concentrations are lower than the referenced data?

C12869
Reply: The PM10 mass concentration in the current study are higher than those re-ported in previous studies mainly due to variation of meteorological parameters especially precipitation event during sampling days.This point will be added in the revised MS.Please see lines 245-247.P11.Line 11: OC/EC ratio from biomass burning and vehicle exhaust are quite different.For example, Saarikoski et al. 2008 (Sources of organic carbon in fine particulate matter in northern European urban air) showed that OC/EC ratio derived from biomass burning was 6.6, Sandradewi et al. 2008 (A study of wood burning and traffic aerosols in an Alpine valley using a multi-wavelength Aethalometer) showed the ratio from wood burning was 7.3, but from vehicle exhaust was 1.1.More discussions need here to prove that high values of OC are from biomass burning.As for the low OC/EC ratios, why is it not from the mixture aerosols of biomass burning and vehicle exhaust?Reply: Our sampling site is a rural site where aerosol contribution from vehicle exhausts is not significant.However, different biomass burning activities could contribute to the variation in OC/EC ratios as suggested by the reviewer.This point is added in the revised MS.Please see lines 281-282.Reply: The reason for different r2 between seasons could indicate that the impacts of local emission on organic aerosol concentrations are significant compared to the influence of long-range transported aerosol (especially in wet season).This point is added in revised MS.Please see lines 285-289.P12.Line 5: There was a significant enhancement of WSOC in Figure 4b between 29/7 and 3/8 as the authors described.Levoglucosan was also elevated during these days in Figure 5.It is better to combine the different compounds to illuminate the sources.
Reply: A new sentence "Similarly, those days with higher WSOC and elevated lev-C12870 els of levoglucosan were observed (Fig. 5), suggesting a possible biomass burning contribution to WSOC" will be added in revised MS.Please see lines 302-305.
P12. Line 10: There is no discussion and no showing in Figure 5 of mannosan in this section, but why the title contains mannosan?At the end of this paragraph, could you distinguish these different burning sources?
Reply: The term of Mannosan will be deleted in the title and the revised text.Detailed results on anhydrosugars will be reported in a manuscript that is in preparation.P13.Line 6: "The ion distributions during the episode (1-3 August) are more pronounced with NH4+ and K+ in PM2.5, in which the abundance of NH4+ was comparable to that of SO42-(Fig.6)."I cannot find K+ in PM2.5 during August 1-3, so why it is pronounced?It seems that the major ions are NH4+, Mg2+, NO3-, Ca2+ in PM2.5.The description in the text and figure showing are not constant.The same situation happens for the PM10.Please check carefully the text and figure 6.
Reply: We thank for the reviewer's comments.The descriptions of the text and Fig. 6 were corrected in the revised MS by including the following sentences: "The temporal variations of ions are shown in Fig. 6.It is noteworthy that daily variability of the ion in PM2.5 and PM10 exhibited similar patterns with higher levels during 1−3 August.The ion distributions during the episode (1−3 August) are more pronounced with NH4+ and NO3− in PM2.5 and Mg2+, Ca2+ and NO3− in PM10 (Fig. 6)."Please see lines 337-341 in the revised MS.
P13. Line17: "low temperature and enhanced biomass burning activities (an important source of NOx) may be responsible for the difference in NO3-levels between the two seasons".In fact there is no temperature difference in the two seasons as you descript in section 2.3.Vehicle exhaust in the city nearby is not a possible reason?
Reply: The reviewer is correct in terms that there was no temperature difference between the two sampling campaigns.The sentence was reworded in revised MS.Please C12871 see lines 334-337.P15.Line 7: "The production of WSOC was slightly more favorable during the wet season", what is the reason?It seems inconsistent with your explanation "SOA is more hydrophilic and one would expect larger WSOC/OC in dry season than in wet season".
Reply: The sentences were reworded and the reason was given.Short-chain polar organic compounds have functional groups that may lead to more water-soluble fraction in WSOC.Please see line 379-381 in the revised MS.P15.Line 22: EC and BC are not directly comparable due to their measurement principles.When you compare the EC/TC ratios with BC/TC in the references, you should remind reader briefly the difference.
Reply: We agree with the reviewers comment.A text describing the difference between EC and BC will be added in lines 388-391 in the revised MS.P19.Line7: It is interesting that nss-K+ showed a stronger correlation with OC in PM10.Could you explain briefly the possible reason?
Reply: The sentence will be reworded in the revised MS by explaining the reason for the correlation of nss-K+ and OC.Please see lines 485-488.P19.Line 19: What is the reason for the higher nss-K+/EC ratios in Tanzania?Is it due to the higher nss-K+ or lower EC compared with other nations?
Reply: In this study we observed high nss-K+ and low EC concentrations.Biomass burning is the main source for nss-K+ at our site.This point will be added in several sentences including lines 468-469 and 495-496 in the revised MS.P19.Line 29: Are there any reference data of PM2.5 nss-SO42-/OC (or EC) and NO3-/OC from biomass burning aerosols which are comparable to your ratios?If yes, then you could conclude the last sentence in this section that Morogoro are more influenced C12872 by biomass burning but not fossil fuel combustion.
Reply: A sentence "On the other hand, our SO42−/OC and NO3−/OC ratios are comparable to those reported from individual fires in Brazil (Ferek et al., 1998) P21.Line 1: The Lev/Man ratios could be used to distinguish the biomass categories with the lowest ratios for soft wood (3-5) and higher ratios for hard wood and crop residues (>10).It is hard to say whether Tanzania has more influence from hard wood and crop residue burning if without calculation or investigation.Even though the burning amount of soft wood is the same as the other two, the Lev/Man ratios could be still > 10.So it is better to express that Tanzania is influenced by mixture aerosols from softwood, hard wood and crop residue burning.
Reply: We agree with reviewer and appreciate the comment.The phrases were reworded in the revised MS in lines 546-549 and 559-562.Reply: We thank the reviewer's comment for providing us with potential references in calculating the biomass burning contributions to OC.However, as replied for the former comments, our study has not be done at near source and emission factors used in other C12873 studies cannot be a representative for our site due to differences in geography, climate and vegetation species between regions.We therefore feel that the levoglucosan to OC ratios should better give a picture of biomass burning contribution to OC.

Anonymous Referee #2
We thank the reviewer#2 for his/her detailed and constructive comments on our manuscript.Below are the point-by-point replies to the comments and suggestions.This manuscript presents data on the PM10 and PM2.5 aerosol composition during wet and dry season campaigns in the year 2011 at Morogoro, Tanzania.The parameters measured were the PM mass, the carbonaceous components OC, EC, and TC (= OC + EC), the major inorganic ionic species (also MSA-), and the anhydrosugars levoglucosan and mannosan.Similar measurements at Morogoro were performed in 2005 and 2006 wet and dry season campaigns (Mkoma et al., 2009a(Mkoma et al., ,b, 2010a,b),b).For the earlier campaigns, also many elements (including important indicators for soil dust) were measured.As a consequence, there is little novel in the present manuscript.
Reply: Thank you for the comments, however we would like to point out the originality of this work as follows: (1) there had been no extensive studies conducted on source implication for organic aerosols in Tanzania, Africa, (2) the study on levoglucosan in aerosols has been conducted in this study for the first time in Tanzania.These points will be added in the revised MS.Please see lines 84-86 and 92-93.Moreover, there is too much repetition and re-iteration of what was already written in the previously published papers of the first author.The present manuscript also suffers from a lack of focus and there is too little data analysis and novel data interpretation in it.The only really worthwhile novel data in the present manuscript are those for lev-C12874 oglucosan and mannosan, but much too little is done with those new data.Therefore, I cannot recommend publication of the manuscript.I suggest that the authors write a new, much more concise manuscript and concentrate in it on the biomass burning indicators levoglucosan, mannosan, and water-soluble non-sea-salt potassium (nss-K+), perhaps complemented with other ionic species, which may have a contribution from biomass burning, and relate the biomass indicators to OC (or even TC).Interspecies ratios of the biomass burning indicators and ratios of the biomass burning indicators to OC (or TC) should be compared with (and related to) literature data for tropical and sub-tropical sites that were (heavily) impacted by biomass burning and with emission factor data given in Andreae and Merlet (2001).Note that Table 1 of the latter paper contains emission factors for levoglucosan, K, OC, and TC for various types of biomass burning, which could be used for obtaining interspecies ratios.By relating these ratios (and also the ones from other appropriate literature references) to their own ratios the authors should attempt to arrive at a much better assessment of the impact from biomass burning to the OC (or TC) and of the relative impact or the various biomass burning sources.
Reply: We appreciate the reviewer's comments.For better clarity and focus on data analysis in the revised MS, we deleted/shorten some sentences particularly in the introduction and result & discussion sections.Please see the revised MS, in which removed sentences/phrases are shown.Although inorganic ions from the Tanzanian aerosols have been reported and discussed in Mkoma et al. (2009a,b;2010a,b), the sampling sites are not the same.Thus we believe that it is worthwhile to discuss the data of this study for the further atmospheric implication in this MS.Unfortunately, our study cannot explain the emission factors for various types of biomass burning.We appreciate the reviewer's comments.However, we feel that those points would be our future research target in the area.Based on the reviewer's comments, however, sentences have been added to compare the various ratios of biomass burning indicators reported by Andreae and Merlet (2001) with our study to explain the contribution of biomass burning to OC. Please see lines 555-562 and 581-584 in the revised MS.We also added K+ C12875 and LG/K+ ratios from our study in Table 2. Please see lines 967.
The IMPROVE thermal protocol (with thermal-optical transmission (TOT) correction) was used to obtain the OC and EC data for the current manuscript, whereas a NIOSH like protocol (also with TOT correction) was used for the previous campaigns at Morogoro (Mkoma et al., 2009a,b, 2010a,b).It is well known that the split between EC and OC in TOT analysis depends on the thermal protocol and that different protocols may provide quite different EC/OC splits, especially for samples that are highly impacted by biomass burning, with the IMPROVE protocol giving larger EC data than the NIOSH protocol.Considering that EC is by far the minor of the two carbonaceous components (OC and EC) in the authors' samples (and is generally the smallest of the two in other sample sets), the impact of the protocol on the EC values will be quite large, whereas the OC data are much less influenced by it (note that the TC data are independent of the protocol).As a consequence, one should be very careful in comparing one's EC data or ratios with EC in the numerator or denominator with data or ratios published in the literature.One should also take care in comparing EC and BC data.I am afraid that the comparisons in lines 10-26 of page 28,679 are not really justified.

P11.
Line 19: What is the reason for the quite different r2 in the wet and dry season (Fig 3a)?

P21.
Line 16: For calculating the biomass burning contributions to OC, the authors could refer to the literatures by Sang et al. 2011 (Levoglucosan enhancement in ambient aerosol during springtime transport events of biomass burning smoke to Southeast China), Zdrahal et al., 2002 (Improved method for quantifying levoglucosan and related monosaccharide anhydrides in atmospheric aerosols and application to samples from urban and tropical locations), Puxbaum et al., 2007 (Levoglucosan levels at background sites in Europe for assessing the impact of biomass combustion on the European aerosol background) and Zhang et al., 2010 (Chemical speciation, transport and contribution of biomass burning smoke to ambient aerosol in Guangzhou, a mega city of China).

Figure 3 :
There is Fig.3bin the text (P11), the labels (a and b) should also be shown in the figure.Reply: Fig3and the text (P11) were modified in the revised MS.
" and references "Gillies et al., 2001 and Ferek et al., 1998" will be added in the revised MS.Please see lines 505-508.