Earthdata Forum

Hello,

I noticed a significant and sudden increase in springtime biomass burning emissions (BCEMBB, OCEMBB, SO2EMBB) over the Indochina Peninsula from 2008 to 2010. The dataset I am analyzing is from MERRA-2 tavgM_2d_adg_Nx: 2d,Monthly mean,Time-averaged,Single-Level,Assimilation,Aerosol Diagnostics (extended) V5.12.4 ( M2TMNXADG) (GES DISC Dataset: MERRA-2 tavgM_2d_adg_Nx: 2d,Monthly mean,Time-averaged,Single-Level,Assimilation,Aerosol Diagnostics (extended) V5.12.4 (M2TMNXADG 5.12.4)).

I have reviewed the 2017 MERRA-2 overview paper (https://journals.ametsoc.org/view/journals/clim/30/17/jcli-d-16-0609.1.xml?rskey=0Swyjp&result=10) and the technical document "The MERRA-2 Aerosol Assimilation (https://gmao.gsfc.nasa.gov/pubs/docs/Randles887.pdf)."

While I understand there is a correction due to different input sources in the dataset, I am unable to find additional information to explain this large jump in the area average timeseries. Any thoughts, comments, or advice would be greatly appreciated!

Mien-Tze

Can you send us the figure where you noticed a sudden increase in springtime biomass-burning emissions? Please also specify the region.

The biomass-burning emission inventory used in MERRA-2 stayed the same in 2009 but changed to QFED in 2010. Please refer to Table 1 in Randles et al. (2017).

I plotted the time series with Giovanni and found no sudden increase. In case you don't know, you can explore MERRA-2 data in Giovanni https://giovanni.gsfc.nasa.gov/giovanni/#service=ArAvTs&starttime=1996-01-01T00:00:00Z&endtime=2015-12-31T23:59:59Z&bbox=92.1094,8.5693,113.9063,25.4443&data=M2TMNXADG_5_12_4_BCEMBB

Here are two MERRA-2 papers for aerosol assimilation for your reference.

Randles, C. A., da Silva, A. M., Buchard, V., Colarco, P. R., Darmenov, A., Govindaraju, R., Smirnov, A., Holben, B., Ferrare, R., Hair, J., Shinozuka, Y., Flynn, C. J.. 2017. The MERRA-2 Aerosol Reanalysis, 1980 Onward. Part I: System Description and Data Assimilation Evaluation. Journal of Climate. Vol. 30, No. 17, pp. 6823-6850. DOI: 10.1175/JCLI-D-16-0609.1 ISSN: 0894-8755, 1520-0442

Buchard, V., Randles, C. A., da Silva, A. M., Darmenov, A., Colarco, P. R., Govindaraju, R., Ferrare, R., Hair, J., Beyersdorf, A. J., Ziemba, L. D., Yu, H.. 2017. The MERRA-2 Aerosol Reanalysis, 1980 Onward. Part II: Evaluation and Case Studies. Journal of Climate. Vol. 30, No. 17, pp. 6851-6872. DOI: 10.1175/JCLI-D-16-0613.1 ISSN: 0894-8755, 1520-0442

Thank you for the reply.
The region is 90-110E, 10-30N. Springtime is March to April.

I have plotted the timeseries and downloaded the data from Giovanni. The result are the same as my previous result. Please find the attached files of the plots.

I do not expect an increasing trend over the Indochina as I did not find any increasing signal from the MODIS fire spot dataset. So, any suggestion?

Thank you for the assistence.

MT

Your plots are helpful. The biomass burning emissions during springtime increased after 2010. There are two possible reasons explaining this sudden increase:

1) The use of QFED emission inventory after 2010. QFED is relatively higher than GFED. You can refer to Pan et al., 2020 for their comparisons and more references for each emission inventory.

2) The emissions from biomass burning increased after 2010. You can check this with another emission inventory, e.g., GFEDv4s.

Here is a plot showing each season's MERRA-2 BC biomass burning emissions. The DJF shows a decreasing trend, which is interesting.
Reference:
Pan X., C. Ichoku, M. Chin, et al. 2020. "Six Global Biomass Burning Emission Datasets: Inter-comparison and Application in one Global Aerosol Model." Atmos. Chem. Phys 20 (2): 969–994 [10.5194/acp-20-969-2020]

Hello,
I am chiming in to confirm that there was a change in the source data for biomass burning emissions in 2010 in MERRA-2. Prior to 2010, emissions were monthly and scaled to GFEDv3.1 on a global scale. Beginning in 2010, emissions are daily from QFEDv2.4r6. The source data can be used to determine whether this is the cause for the change in character. It may be worthwhile to plot the same timeseries over Indochina solely for QFED to confirm if the change in dataset influenced the emissions. QFEDv2.4r6 data beginning in 2000 can be downloaded from https://portal.nccs.nasa.gov/datashare/iesa/aerosol/emissions/QFED/v2.4r6/.

Allie

Thank you very much for your response.

The reference for dataset inter-comparison and the plot showing a decreasing trend of winter emissions are very helpful. I will review the reference thoroughly and shall conduct further research on the DJF decreasing trend (very interesting).

I appreciate your assistance. Have a great day.

MT

Hi Allie,

I think it can be worthwhile to have a try. I will do some research on QFEDv2.4r6 data. Thanks!

MT

This is a follow-up to my previous post regarding the increase in springtime biomass burning emissions after 2010. In addition to QFEDv2.4r6, I have also examined other biomass burning emission inventories, including QFEDv2.6r1, GFED4s, GFED5, FINNv2.5, and FEER. Here, I present two plots of the area-averaged BC biomass burning emissions: seasonal means for each of these emission inventories, and a comparison of monthly means (February, March, and April) among these inventories.

My brief notes on these plots:
1) The sudden increase in springtime biomass burning emissions after 2010 is primarily due to a change in the source data, from monthly GFED prior to 2010 to daily QFED starting in 2010.
2) The DJF decreasing trend present in the MERRA-2 biomass burning emissions, where February is the major contributor, is likely a real one.

I don't have much experience in biomass burning emissions and am unable to interpret these plots further. I hope this updated post provides more information for anyone who may read it in the future. I plan to look at the long-term variability of MERRA-2 AOD over Indochina later, although I am not quite sure what to expect at this time. Any thoughts, comments, or advice would be greatly appreciated!

- Data sets:
- QFEDv2.4r6 https://portal.nccs.nasa.gov/datashare/iesa/aerosol/emissions/QFED/v2.4r6/
- QFEDv2.6r1 https://portal.nccs.nasa.gov/datashare/iesa/aerosol/emissions/QFED/v2.6r1/
- GFED4s https://www.geo.vu.nl/~gwerf/GFED/GFED4/
- GFED5 https://www.globalfiredata.org/data.html
- FINN v2.5 https://www2.acom.ucar.edu/modeling/finn-fire-inventory-ncar
- FEER https://feer.gsfc.nasa.gov/data/emissions/

Thanks for your thorough comparisons with various biomass-burning emission inventories! I agree with most of your conclusions with additional notes.

1) The sudden increase in springtime MAM biomass burning emissions in MERRA-2 after 2010 is spurious and absent in other biomass burning inventories. This increase is primarily due to an artificial change in the source data, from monthly GFED before 2010 to daily QFED starting in and after 2010, especially in March and April (I am still determining about May).

2) The DJF decreasing trends are present in the MERRA-2 biomass-burning emissions and other biomass-burning inventories. Is February the major contributor? I will be unsure until I see the December and January plots, as you did for February, March, and April.

Xiaohua

Xiaohua,

Yes, I did have the plots for other months. Here, I present two plots of the area-averaged BC biomass burning emissions: monthly means for each of these emission inventories, and a comparison of monthly means among these inventories. The format is similar to my previous post for your reference.

1) 12 months for each of these emission inventories: For MAM, the BC emissions in May are the lowest in MERRA-2 and all biomass burning inventories. For DJF, February is the major contributor, except in GFED4s and FEER, where January is comparable to February.
2) 12 months for inter-comparison among these emission inventories: Both December and January show a period of larger emissions during the 2000s, yet the contribution of February to the DJF decreasing trend is clear. The variability of May shows a somewhat different pattern.
3) From the plots of inter-comparison, another interesting feature can be observed: The effect of the artificial change in the source data can be seen from March to April, and also from June to October. For the rest, it is hard to tell, and among them, only May is in the warm season.

**Note: In the plots for inter-comparison, for plotting purposes, FEER was multiplied by 0.1 for December, January, and May, and multiplied by 0.01 for months from June to November.