MODIS Aerosol (MOD04·MOD08)¶

MODIS/Terra and Aqua Aerosol 5-Min L2 Swath 10km and 3km (MOD04_L2, MYD04_L2, MOD04_3K, MYD04_3K) and MOD08/MYD08 L3 1° aerosol, Collection 6.1 (MODIS Dark Target and Deep Blue aerosol optical depth)

Being replaced

At a glance¶

Note on period: Terra (MOD04, MOD08) from 2000-02-24; Aqua (MYD04, MYD08) from 2002-07-04. On 2026-10-13 LAADS held data up to early October 2026 (MOD08_D3 A2026282). C6.1 was reprocessed for the full record from September to December 2017 (Terra) and January to March 2018 (Aqua); since 30 April 2018 only C6.1 has been produced. No new data will appear once Terra and Aqua operations end.

Overview¶

Clear-sky aerosol optical depth (AOD) retrieved from MODIS visible and near-infrared reflectance. Each file carries two families of retrievals from different algorithms. Dark Target (DT) works over ocean and over dark, vegetated land; Deep Blue (DB) works over all land, including bright desert surfaces; a combined AOD (AOD_550_Dark_Target_Deep_Blue_Combined) merges the two. Collection 6.1 improved DT retrievals over urban areas and added per-pixel uncertainty estimates to DB. The products are the standard 10 km L2, a 3 km DT L2 for air quality work and 1° gridded L3 (daily, 8-day, monthly), with separate Terra (morning) and Aqua (afternoon) versions.

What changes with the successor¶

MODIS aerosol C6 (MOD04, MOD08) → MODIS aerosol C6.1¶

Transition: 2018-03

A version that fixed problems in the input L1B data and added some L2 and L3 improvements. The full record was reprocessed from September to December 2017 for Terra (public release 15 October 2017) and January to March 2018 for Aqua; since 30 April 2018 only C6.1 has been produced. DT gained improved retrievals over urban areas and DB gained per-pixel uncertainty estimates.

Sources: MODIS Atmosphere — Data Versions (collections overview) · C6.1 Dark Target change document · C6.1 Deep Blue change document

MODIS aerosol MOD04, MOD08 C6.1 → VIIRS Deep Blue AERDB and Dark Target AERDT Planned¶

The successors are continuity products that apply the same Deep Blue and Dark Target algorithms to VIIRS (Suomi NPP, NOAA-20, NOAA-21). The two algorithms become separate products (AERDB and AERDT); DB gains an ocean algorithm (SOAR) and covers land and ocean, and there is no combined AOD and no 8-day L3. They mainly continue Aqua, which flies in the same afternoon orbit; no VIIRS continues Terra's morning observations. The transition completes when Terra and Aqua end operations in 2027.

Sources: Sawyer et al. (2020), Remote Sensing — Continuing the MODIS Dark Target Aerosol Time Series with VIIRS · Sawyer et al. (2025), Frontiers in Environmental Science — updated VIIRS DT and MODIS continuity · Sayer et al. (2019), JGR — consistency of MODIS C6.1 and VIIRS Deep Blue · LAADS DAAC — MODIS to VIIRS Transition

Compare the two side by side

Available variables¶

SDS names from the CMR variable records (UMM-Var) and the C6.1 documents. Files are HDF-EOS2 (HDF4) with integer storage, so physical value = scale_factor × (stored value − add_offset), as defined in the MODIS Atmosphere L3 ATBD (add_offset is subtracted). AOD is dimensionless and referenced to 550 nm (0.55 µm). MYD (Aqua) files use the same names.

instantaneous mean accumulated maximum minimum static — instantaneous: value at the time stamp · mean/accumulated/maximum/minimum: over the indicated interval · static: does not change with time

L2 swath 10 km (MOD04_L2, MYD04_L2) and 3 km (MOD04_3K, DT only)¶

L3 daily 1° (MOD08_D3, MYD08_D3)¶

L3 8-day and monthly 1° (MOD08_E3 8-day / MOD08_M3 monthly)¶

The full variable list is in the official documentation.

How to get the data¶

• LAADS DAAC (search, order and direct download) — Web search, direct HTTPS download (archive/allData/61//) / LAADS API-V2 / OPeNDAP
  • Example short names: MOD04_L2, MYD04_L2, MOD04_3K, MYD04_3K, MOD08_D3, MOD08_E3, MOD08_M3 (MYD for Aqua). All are C6.1 (archive 61; CMR version "6.1"). Since 2026-04-30 OPeNDAP also requires Earthdata Login, and the old LAADS Classic API was retired on 2026-08-31, so scripts must move to API-V2.
• NASA Earthdata Search — Web search / earthaccess (Python)
  • The link points to MOD08_D3 C6.1; search other products by short name. Files are also served from Earthdata Cloud (s3://prod-lads).
• LANCE near-real-time (NRT) MOD04_L2 — Web search / HTTPS
  • Near-real-time version (version 6.1NRT, DOI 10.5067/MODIS/MOD04_L2.NRT.061, from 2017-10-11). Use it when speed matters, such as hazard monitoring; use the standard product for research time series.

Example code¶

Downloads a Terra MOD08_D3 daily file with earthaccess and reads the combined AOD daily mean with pyhdf. MODIS atmosphere products subtract add_offset from the stored value before multiplying by scale_factor, so convert by hand.

# pip install earthaccess pyhdf numpy
import earthaccess
import numpy as np
from pyhdf.SD import SD, SDC

earthaccess.login()  # free NASA Earthdata account
results = earthaccess.search_data(
    short_name="MOD08_D3", version="6.1",
    temporal=("2024-04-01", "2024-04-03"),
)
files = earthaccess.download(results, "mod08")

hdf = SD(files[0], SDC.READ)
sds = hdf.select("AOD_550_Dark_Target_Deep_Blue_Combined_Mean")
raw = sds.get().astype("float64")
att = sds.attributes()
aod = np.where(raw == att["_FillValue"], np.nan,
               att["scale_factor"] * (raw - att["add_offset"]))  # 1° grid, 180×360
# row 0 = 89–90°N, column 0 = 180–179°W
lat = 89.5 - np.arange(180); lon = -179.5 + np.arange(360)
j = np.argmin(abs(lat - 37.5)); i = np.argmin(abs(lon - 127.0))
print("AOD in the 1° cell near Seoul:", aod[j, i])

Practical tips¶

• Which algorithm feeds the combined AOD (AOD_550_Dark_Target_Deep_Blue_Combined) changes with region and season (see _Algorithm_Flag). In a long record for one region, switching algorithms can show up as a jump in the values, so use DT or DB alone if needed.
• For quantitative work use Optical_Depth_Land_And_Ocean (ocean QA ≥ 1, land QA = 3) or the DB _Best_Estimate (QA 2 and 3); do not use all-quality fields such as Image_Optical_Depth_Land_And_Ocean.
• In the HDF4 files the physical value is scale_factor × (stored value − add_offset). The sign is opposite to the common 'stored × scale + offset' convention, so take care when reading with pyhdf and similar tools. Values outside valid_range are not screened out.
• Terra and Aqua differ in overpass time and calibration, so their AOD levels differ slightly (Sawyer et al., 2025 report DT Terra about 0.015 higher than Aqua). Rather than averaging both together, pick one satellite as the reference, and check the overpass time change in the orbit drift period after 2020 (Terra) and 2021 (Aqua).
• The 3 km product (MOD04_3K) uses the same DT algorithm and lookup tables but changes pixel selection, the minimum number of good pixels and quality handling for air quality users. It has DT values only, with no combined AOD or DB fields.
• According to the MODIS Atmosphere issues list, the C6.1 DB cloud screening update that reduces low stratus contamination applies only to data produced since autumn 2019; earlier data will get it in the next reprocessing. Be careful when comparing DB records over marine stratus regions across that date.
• L3 daily means (_Mean) collect only clear pixels, so cloudy days and regions have few samples. Check _Pixel_Counts, and interpret monthly cells with few valid days with care.
• AOD is not the same thing as surface PM2.5. It is column-integrated extinction, so its relationship with surface concentration depends on boundary layer height and humidity.

Notices and change history¶

References for your paper¶

Dataset citation¶

MOD04_L2 C6.1 (Terra 10 km L2 aerosol) — DOI 10.5067/MODIS/MOD04_L2.061

@misc{mod04_l2_c61,
  author    = {Levy, R. and Hsu, C. and others},
  title     = {{MODIS} Atmosphere {L2} Aerosol Product},
  year      = {2015},
  publisher = {NASA MODIS Adaptive Processing System, Goddard Space Flight Center, USA},
  doi       = {10.5067/MODIS/MOD04_L2.061}
}

MYD04_L2 C6.1 (Aqua 10 km L2 aerosol) — DOI 10.5067/MODIS/MYD04_L2.061

@misc{myd04_l2_c61,
  author    = {Levy, R. and Hsu, C. and others},
  title     = {{MODIS} Atmosphere {L2} Aerosol Product},
  year      = {2015},
  publisher = {NASA MODIS Adaptive Processing System, Goddard Space Flight Center, USA},
  doi       = {10.5067/MODIS/MYD04_L2.061}
}

MOD08_D3 C6.1 (Terra L3 daily) — DOI 10.5067/MODIS/MOD08_D3.061

@misc{mod08_d3_c61,
  author    = {Platnick, S. and others},
  title     = {{MODIS} Atmosphere {L3} Daily Product},
  year      = {2015},
  publisher = {NASA MODIS Adaptive Processing System, Goddard Space Flight Center, USA},
  doi       = {10.5067/MODIS/MOD08_D3.061}
}

MOD08_M3 C6.1 (Terra L3 monthly) — DOI 10.5067/MODIS/MOD08_M3.061

@misc{mod08_m3_c61,
  author    = {Platnick, S. and others},
  title     = {{MODIS} Atmosphere {L3} Monthly Product},
  year      = {2015},
  publisher = {NASA MODIS Adaptive Processing System, Goddard Space Flight Center, USA},
  doi       = {10.5067/MODIS/MOD08_M3.061}
}

Reference papers¶

MODIS Collection 6 aerosol (Levy et al., 2013) — DOI 10.5194/amt-6-2989-2013

@article{levy2013c6,
  author  = {Levy, R. C. and Mattoo, S. and Munchak, L. A. and Remer, L. A. and Sayer, A. M. and Patadia, F. and Hsu, N. C.},
  title   = {The Collection 6 {MODIS} aerosol products over land and ocean},
  journal = {Atmospheric Measurement Techniques},
  volume  = {6},
  number  = {11},
  pages   = {2989--3034},
  year    = {2013},
  doi     = {10.5194/amt-6-2989-2013}
}

DT, DB and merged AOD compared, with usage recommendations (Sayer et al., 2014) — DOI 10.1002/2014JD022453

@article{sayer2014merged,
  author  = {Sayer, A. M. and Munchak, L. A. and Hsu, N. C. and Levy, R. C. and Bettenhausen, C. and Jeong, M.-J.},
  title   = {{MODIS} {Collection 6} aerosol products: Comparison between {Aqua}'s e-{Deep Blue}, {Dark Target}, and ``merged'' data sets, and usage recommendations},
  journal = {Journal of Geophysical Research: Atmospheres},
  volume  = {119},
  pages   = {13965--13989},
  year    = {2014},
  doi     = {10.1002/2014JD022453}
}

Validation and consistency of MODIS C6.1 and VIIRS Deep Blue (Sayer et al., 2019) — DOI 10.1029/2018JD029598

@article{sayer2019dbval,
  author  = {Sayer, A. M. and Hsu, N. C. and Lee, J. and Kim, W. V. and Dutcher, S. T.},
  title   = {Validation, stability, and consistency of {MODIS} {Collection 6.1} and {VIIRS} {Version 1} {Deep Blue} aerosol data over land},
  journal = {Journal of Geophysical Research: Atmospheres},
  volume  = {124},
  pages   = {4658--4688},
  year    = {2019},
  doi     = {10.1029/2018JD029598}
}

Acknowledgements¶

Cite the LAADS dataset DOI of each product you use (pattern 10.5067/MODIS/.061, e.g. 10.5067/MODIS/MOD04_3K.061 for the 3 km product and 10.5067/MODIS/MOD08_E3.061 for the 8-day product). Cite Levy et al. (2013) for DT, and Sayer et al. (2014) and Sayer et al. (2019) for DB and the combined AOD, as the algorithm references.

Information on the datasets used in this study was compiled with the help of the Korean Climate Data Navigator (KCDN) (Sim, 2026; https://doi.org/10.5281/zenodo.23300923).

Links¶

• MOD04_L2 product page (LAADS DAAC)
• MOD08_D3 product page (LAADS DAAC)
• MODIS Atmosphere — Aerosol (04_L2) product description
• MODIS Atmosphere — Aerosol known issues (C6.1)
• MODIS Atmosphere L3 User's Guide & ATBD (C6 and C6.1)
• MOD04 ATBD
• Dark Target algorithm team
• LAADS DAAC — MODIS to VIIRS Transition