Earthdata Forum

Answer:
Provided below are recommendations for using MERRA-2 ozone products at various vertical levels. Refer to Wargan et al. (2017) for more information.

• Total column ozone: use the total ozone in inst1_2d_asm_Nx.

• Below 500 hPa, including at the surface, ozone loses realism because the model does not include surface emissions and NOx chemistry, and no vertically resolved observations are assimilated in the lower troposphere. In particular, surface and the lowermost tropospheric ozone values from MERRA-2 are not recommended for science. MERRA-2 ozone has been evaluated between the middle troposphere (500 hPa) and 1 hPa. It is suitable for science within that layer, with caveats delineated in Wargan et al., 2017.

• When using upper stratospheric ozone (at pressures less than ~5 hPa), it is crucial to exercise caution. The assimilation has little impact as chemical timescales for ozone become shorter than the frequency of assimilated data. Additionally, in the Solar Backscatter Ultraviolet SBUV period (1980-2004), the observation errors are set to large values in the upper stratosphere, further limiting the impact of assimilation.

• Mesospheric ozone (pressures less than ~1 hPa) in MERRA-2 is NOT scientifically meaningful, and should NOT be used.

When integrating ozone profiles vertically for a project, the recommendation is to use the native vertical coordinate collections inst3_3d_chm_Nv (or M2I3NVCHM), or inst3_3d_asm_Nv (or M2I3NVASM). This approach offers greater reliability compared to using the pressure levels collections, inst3_3d_asm_Np (or M2I3NPASM) and inst6_3d_ana_Np (or M2I6NPANA), ensuring the accuracy of these data.

For reference, listed below are MERRA-2 variables related to ozone, available in various temporal resolutions (hourly, 3-hourly, 6-hourly, and monthly mean) and vertical resolutions (pressure levels and model levels). Note that TO3 (total column ozone) is present in two 2D hourly mean collections below, i.e., 1) tavg1_2d_chm_Nx and 2) tavg1_2d_slv_Nx, and O3 (ozone mass mixing ratio) in two 3D 3-hourly collections, 6) inst3_3d_chm_Nv, and 7) inst3_3d_asm_Nv. The CHM collections 1) and 6) were conservatively remapped (simply a binning routine) from the native cube sphere grid to the regular latitude-longitude grids. In contrast, the SLV (2) and ASM (7) collections are non-conservatively interpolated using bilinear interpolation. As a rule of thumb, only the data conservatively remapped will balance to the highest precision. For all collections, please use this link to access a specific collection.

MERRA-2 provides the ozone collections below:
2D collections
1) Hourly mean: tavg1_2d_chm_Nx (or M2T1NXCHM): Carbon Monoxide and Ozone Diagnostics. Note: its monthly mean collection is tavgM_2d_chm_Nx (or M2TMNXCHM).
TO3 (total column ozone). Units: DU
2) Hourly mean tavg1_2d_slv_Nx (or M2T1NXSLV): Single-Level Diagnostics. Note: its monthly mean collection is tavgM_2d_slv_Nx (or M2TMNXSLV).
TO3 (total column ozone). Units: DU
3) Hourly instantaneous: inst1_2d_asm_Nx (or M2I1NXASM): Single-Level Diagnostics. Note: its monthly mean collection is instM_2d_asm_Nx (or M2IMNXASM).
TO3 (total column ozone). Units: DU
3D collections
42 Pressure levels
4) inst3_3d_asm_Np (M2I3NPASM): Assimilated Meteorological Fields. Note: its monthly mean collection is instM_3d_asm_Np (or M2IMNPASM).

• O3 (ozone mass mixing ratio). Units: kg kg-1

5) inst6_3d_ana_Np (M2I6NPANA): Analyzed Meteorological Fields. Note: its monthly mean collection is instM_3d_ana_Np (or M2IMNPANA).

• O3 (ozone mass mixing ratio). Units: kg kg-1

72 Model levels
6) inst3_3d_chm_Nv (or M2I3NVCHM): Carbon Monoxide and Ozone Mixing Ratio. Note: there is no monthly mean collection for this case.

• O3 (ozone mass mixing ratio). Units: kg kg-1

7) inst3_3d_asm_Nv (M2I3NVASM): Assimilated Meteorological Fields. Note: there is no monthly mean collection for this case.

• O3 (ozone mass mixing ratio). Units: kg kg-1

In addition, MERRA-2 also provides ozone diagnostics collections below:
2D collections
8) Hourly mean: tavg1_2d_int_Nx (or M2T1NXINT): Vertically Integrated Diagnostics

• DOXDT_ANA (vertically integrated ozone tendency due to analysis). Units: kg m-2 s-1

• DOXDT_DYN (vertically integrated ozone tendency due to dynamics). Units: kg m-2 s-1

3D collections
9) 3 hourly mean: tavg3_3d_odt_Np (or M2T3NPODT): Ozone Tendencies

• DOXDTANA (total ozone analysis tendency). Units: mol mol-1 s-1

• DOXDTDYN (tendency of ozone due to dynamics). Units: kg kg-1 s-1

The model uses an odd oxygen mixing ratio as its prognostic variable. Its vertically integrated tendency is given by:

• DOXDT = DOXDT_DYN + DOXDT_PHY + DOXDT_ANA.

The dynamics contribution is simply the vertically integrated convergence. The physics contributions result from parameterized production and loss terms, mostly in the stratosphere. In MERRA-2, ozone is analyzed and so can change due to ozone increments or due to increments in atmospheric dry mass. Only the total analysis contribution is reported. In order to compare the tendencies with the total tendency from the states, a conversion factor of 1.655 mol/mol must be applied to the tendencies.

For additional information on MERRA-2 data, please read the MERRA-2 File Specification Document first, and refer to the MERRA-2 project page.

In addition, the ozone product in the MERRA-2 Stratospheric Composition Reanalysis of Aura Microwave Limb Sounder (M2-SCREAM) collection is recommended.
https://disc.gsfc.nasa.gov/datasets?keywords=m2-scream&sort=dataset&page=1. More details about M2-SCREAM can be found in Wargan et al. (2023).

References:
Wargan, K., G. J. Labow, S. M. Frith, S. Pawson, N. J. Livesey, and G. S. Partyka (2017) Evaluation of the ozone fields in NASA's MERRA-2 reanalysis. J. Clim., 30, 2961–2988. doi:10.1175/JCLI-D-16-0699.1.
Wargan, K., Weir, B., Manney, G. L., Cohn, S. E., Knowland, K. E., Wales, P. A., & Livesey, N. J. (2023) M2-SCREAM: A stratospheric composition reanalysis of Aura MLS data with MERRA-2 transport. Earth and Space Science, 10, e2022EA002632. doi:10.1029/2022EA002632

Related Data Collections:
M2TMNXCHM_5.12.4
M2T1NXCHM_5.12.4
M2T1NXSLV_5.12.4
M2TMNXSLV_5.12.4
M2I1NXASM_5.12.4
M2IMNXASM_5.12.4
M2I3NPASM_5.12.4
M2IMNPASM_5.12.4
M2I6NPANA_5.12.4
M2IMNPANA_5.12.4
M2I3NVCHM_5.12.4
M2I3NVASM_5.12.4
M2T1NXINT_5.12.4
M2T3NPODT_5.12.4