Skip to main content

NRT Value-Added MODIS Aerosol Optical Depth Product Available

The Naval Research Laboratory and the University of North Dakota developed a level 3 gridded product specifically designed for quantitative applications including data assimilation and model validation. It is available through LANCE-MODIS.
8 MIN READ
Minnie Wong
News
Facebook

Overview

The U.S. Naval Research Laboratory (NRL) and the University of North Dakota (UND) developed a value-added aerosol optical depth dataset based on Moderate Resolution Imagiing Spectroradiometer (MODIS) Level 2 aerosol products. The NRL-UND product is a level 3 gridded product specifically designed for quantitative applications including data assimilation and model validation. This dataset has been given the ESDT identifier MxDAODHD. It is available through the Land, Atmosphere Near real-time Capabilities for EOS-Moderate Resolution Imaging Spectroradiometer (LANCE MODIS). It offers several enhancements over the MODIS Level 2 data on which it is based:

  • Stringent filtering to reduce outliers, eliminate cloud contamination, and exclude conditions where aerosol detection is likely to be inaccurate
  • Reduction of systematic biases over land and ocean by empirical corrections
  • Reduction of random variation in AOD values by spatial averaging
  • Quantitative estimation of uncertainty for each AOD data point

The processing chain to generate this product is conceptually illustrated in the figure below. Calibrated MODIS TOA radiances in visible and near-infrared wavelengths are used as input to the MODIS Level 2 aerosol retrieval algorithms. The Level 2 aerosol data are then used as input, along with several ancillary data sets, to produce the MxDAODHD product. The level 2 aerosol data products are described in publications by Remer et al (2005) and Levy et al. (2007), and the Algorithm Theoretical Basis Document (ATBD) for this product can be found here: http://modis-atmos.gsfc.nasa.gov/MOD04_L2/atbd.html. The filtering and correction algorithms applied to this product are described in publications by Zhang et al. (2006, 2008), Shi et al. (2011), and Hyer et al. (2011) (see “References”, below).

image

Production Details

The MxDAODHD granules are produced every six hours, and time-stamped 00:00, 06:00, 12:00, and 18:00 (all times UTC). Each granule includes MODIS observations from +/-3 hours from the timestamp (e.g. 12:00 product includes MODIS data from 09:00-15:00 UTC). MCDAODHD granules include data from both Terra and Aqua (see “Product Format,” below); MODAODHD and MYDAODHD granules are also separately produced. Production is initiated as soon as the Level 2 inputs become available in the LANCE system. The MxDAODHD process uses MOD04_L2 data as its only dynamic MODIS input. The GEOS-5 meteorology is also used to calculate an empirical wind speed correction. If the GEOS-5 analysis is unavailable, forecast data are used. The MCDAODHD process also uses several static datasets, including a climatological surface albedo derived from the MCD43C3 MODIS albedo product. For details, see the published references listed below. Note that the MOD04_L2 input for this product is the near-real-time product from LANCE, which can exhibit some differences from the final science-grade MOD04_L2 produced by MODAPS.

Product Format and Filespec

The MxDAODHD granules are distributed in HDF-EOS format. The filename is described by MxDAODHD.AYYYYDDD.HHMM.VVV.NRT.hdf, where YYYY is the year, DDD is the day of year (January 1 == 1), HHMM is the hour and minute, and VVV is the MODIS Version (5.1 for the initial release of these data, see FAQ). Product using data from MODIS-Terra only have filenames beginning with MOD, MODIS-Aqua only have filenames beginning with MYD, and MCD files use data from both MODIS sensors. The HDF-EOS format permits storage of multiple datasets as swaths, with data attributes tied to geolocation information in the form of a list of observation locations and times. HDF-EOS permits multiple swaths to be stored in a single file. A granule of MCDAODHD includes the following swaths:

  • Ocean_Terra_AOD
  • Ocean_Aqua_AOD
  • Land_Terra_AOD
  • Land_Aqua_AOD
  • Landocean_Terraaqua_AOD

Each of these swaths has a dimension “Nobs,” with a single record for each valid observation in the swath. Each swath contains the following fields:

  • Geolocation fields:
    • Longitude: -180 to 180
    • Latitude: -90 to 90
    • Time: Days since January 1, 4713BC, 12:00
      • Example: 1 October 2012 00:00Z translates to 2456201.5
  • Data fields (all in units of AOD, unless otherwise specified):
    • Aerosol Optical Depth: AOD
    • AOD_Error_SD: estimated total uncertainty of AOD, including observation error and representativeness error, see peer-reviewed publications for details
    • AOD_Grid_Cell_SD: standard deviation of Level 2 AOD retrievals within grid cell
    • AOD_Observation_Error: empirical estimate of AOD observation uncertainty, see peer-reviewed publications for details
    • AOD_N10km_In_Grid: number of Level 2 AOD retrievals within grid cell

Frequently Asked Questions

Q: What is this product?

A. Filtered, Corrected, Gridded Aerosol Optical Depth with estimated error for each data point Q: What scientific value is added to MOD04_L2?

A. This product modifies the MOD04_L2 input data in several ways to improve the utility of these data for quantitative applications. These modifications are described in detail in the publications (see “References,” below), but are briefly described here:

  1. Reduction of outliers by filtering. In addition to filtering using the built-in diagnostics in the MOD04 product, textural and “buddy-check” filters are employed to avoid residual cloud contamination, and problematic land surfaces are avoided using a climatological database of surface spectral albedo.
  2. Reduction of systematic biases by empirical corrections. For over-ocean data, two corrections are employed. For low AOD, the correction is based on wind speed, glint angle, and MODIS retrieved cloud fraction. For high AOD, the applied correction is a function of cloud fraction and MODIS retrieved aerosol fine mode fraction. For over-land data, the low-AOD correction is based on climatological spectral surface albedo, and the high-AOD correction is calculated separately for different regions.
  3. Reduction of random error by averaging. The MOD04 retrievals with a nominal nadir resolution of 10x10km (expanding to roughly 40x20km at the edge of the MODIS swath) are binned to a lat/lon grid with a resolution of one-half degree.
  4. Quantitative uncertainty estimation for each data point. Uncertainty is estimated as the geometric sum of a representativeness term based on variability of L2 retrievals within a grid cell and an observation error term estimated empirically.

Q: Is the product easier to use than MOD04_L2?

A. The product is lightweight (1 file every six hours), and includes only information needed for assimilation. Q: Who needs aerosol data assimilation?

A. Anyone doing aerosol forecasting:

  • Currently operational: Navy FNMOC, NOAA (NOAA GFS Aerosol component currently includes dust aerosol only)
  • Quasi-operational: ECMWF
  • Quasi-operational: NASA GMAO
  • Experimental: University research groups
  • Especially valuable for supporting field/aircraft campaigns

Q: What value does this product have beyond data assimilation?

A. This product has numerous other practical and scientific applications. It has been used for:

  • Quantification of bias in SST retrievals;
  • Cross-comparison of MODIS and other satellite AOD products;
  • Evaluation of sampling bias in regional radiative forcing estimates;
  • Evaluation of 24-72 hour aerosol model forecasts;

Q: Will this product still be relevant once Collection 6 is available?

A. Yes. Collection 6 will provide a higher-quality input, and some of the error corrections developed for use in MCDAODHD are included in the Level 2 product for Collection 6. However, even with those corrections now made upstream, there is value in the noise reduction and error estimates provided by this product.

A2. Changes to the MOD04_L2 product will require downstream changes to the assimilation-ready product. NRL and UND have been through this with C4 to C5. Q: What are the differences between the MODAODHD, MYDAODHD, and MCDAODHD products? Which one should I use?

A. Terra and Aqua products are produced separately, and can be found in the MODAODHD and MYDAODHD granules respectively. The combined product is produced only when both Terra and Aqua processing is complete. The MCDAODHD granules contain the combined data as well as the separate outputs for Terra-land, Terra-ocean, Aqua-land, and Aqua-ocean. Because Terra and Aqua data can have different latencies in the LANCE system, the MOD and MYD products can sometimes be available up to 60 minutes before the MCD product. Also, Terra will pass over a given point on the earth earlier than Aqua, nominally 2 hours earlier at the equator, less towards the North Pole and more towards the South Pole.The MODAODHD and MYDAODHD products are distributed for users with very stringent latency requirements. Q: Is there available example code for reading the MxDAODHD granules?

A. Example code in the IDL language is available here: mcdaodhd_read.pro Q: Is the source code for this product available?

A. Yes. NASA makes the source code for all MODAPS products available to registered users who agree to a standard NASA Software Usage Agreement. In order to register for MODIS software, you can go to the following URL: https://modaps.modaps.eosdis.nasa.gov/ . Once you register, you will need to print and sign a copy of the software agreement (https://modaps.modaps.eosdis.nasa.gov/MODIS_SUA.pdf) and fax the form back to us at Fax # 301-557-6622. After your fax is received, you will be added to the software access group, and will receive an email notification with the URL for the software downloads.

References

Hyer, E. J., J. S. Reid, and J. Zhang (2011), An over-land aerosol optical depth data set for data assimilation by filtering, correction, and aggregation of MODIS Collection 5 optical depth retrievals, Atmospheric Measurement Techniques, 4, 379-408, doi: 10.5194/amt-4-379-2011.

Levy, R. C., L. A. Remer, S. Mattoo, E. F. Vermote, and Y. J. Kaufman (2007), Second-generation operational algorithm: Retrieval of aerosol properties over land from inversion of Moderate Resolution Imaging Spectroradiometer spectral reflectance, J. Geophys. Res.-Atmos., 112(D13), D13211.

Remer, L. A., et al. (2005), The MODIS aerosol algorithm, products, and validation, J. Atmos. Sci., 62(4), 947-973.

Shi, Y., Zhang, J., Reid, J. S., Holben, B., Hyer, E. J., and Curtis, C.: An analysis of the collection 5 modis over-ocean aerosol optical depth product for its implication in aerosol assimilation, Atmos. Chem. Phys., 11, 557-565, doi: 10.5194/acp-11-557-2011, 2011.

Zhang, J. L., and Reid, J. S.: MODIS aerosol product analysis for data assimilation: Assessment of over-ocean level 2 aerosol optical thickness retrievals, J. Geophys. Res.-Atmos., 111, 2006.

Zhang, J. L., Reid, J. S., Westphal, D. L., Baker, N., and Hyer, E. J.: A System for Operational Aerosol Optical Depth Data Assimilation over Global Oceans, Journal of Geophysical Research, 113, doi: 10.1029/2007JD009065, 2008.

Details

Last Updated

Published