Frequently Asked Questions

Last Update: June 22, 2007

This is a transitional FAQ for MODIS data users, particularly those interested in MODIS radiance, geolocation, cloud mask and other atmospheric products. We expect this document to evolve rapidly in response to your feedback and questions.

The goal of this document is to address questions about using the new Land and Atmosphere Archive and Distribution System and to direct researchers to the multitude of MODIS resources that are available.

Q: What is LAADS?

A: LAADS is an online archive of MODIS data products. Products are currently available via FTP, direct web download or web order. The FTP portal is via FTP to machine "ladssci.nascom.nasa.gov." Web ordering supports product selection, subsetting and special processing. Get more details at the LAADS Help Overview.

Q: What MODIS data are available via LAADS?

A: LAADS is being populated with Aqua radiance, geolocation and atmospheric products (denoted "Collection 5") as they are produced. Some earlier "Collection 4" products such as Aqua/Terra Atmosphere Level 3 are available. New Terra products will be included soon. To get an up to date picture, please consult the LAADS Web Data Availability Chart. If your Atmospheric product is not yet available in LAADS, it can still be ordered (in "Collection 4") from the GES-DAAC.

Q: Will I have immediate online access to all of these MODIS products?

A: Yes and no (sorry). Geolocation and Atmosphere products make up the LAADS permanent archive. They remain online and accessible once created. Radiance products are not stored permanently. They are available as a rolling datapool archive or, after deletion, are recreated via web order.

Q: What if I want a lot of MODIS data or a time series of products?

A: Please contact us if you would like to arrange a data subscription or have special requirements. Alternately, some researchers know exactly what they want and use ftp scripts to pull these products from the LAADS Web FTP Site.

Q: Can you start with the big picture and explain this Level-1, Level-2, Level-3 business?

A: Most researchers want retrieved geophysical parameters such as aerosol or cloud properties. These parameters are stored as 5 minute orbital swath slices (MODIS granules) in multi-parameter Atmosphere L2 products. L2 products contain the highest spatial resolution parameters available, typically 1km, 5km or 10k at satellite nadir. Atmospheric L3 products contain global statistical summaries of the L2 swath data. The data are binned in a linear lat/long grid of 1 degree spatial resolution. The L3 summaries are done for daily, 8-day and monthly time periods. The L3 Land products are quite different.

Q: And what about Level-1? Would I be interested?

A: All Land and Atmosphere products are derived from L1B calibrated radiances (at 250m, 500m or 1km resolution) using L1 geolocation and the MODIS cloud mask. These products all span the same 5 minute orbital granules as the L2 products. Some researchers do want radiance data for various purposes, including making great pictures. They may also use the need 1km geolocation files (MOD03/MYD03) in conjunction with the L2 products. Atmosphere L2 products do contain lower resolution lat/long parameters, so it may not be necessary to acquire the L1 geolocation files.

Q: Where can I get technical details about the L1B radiance products?

A: Start with the MODIS Technical Specs page for a complete listing of MODIS spectral bands. The MODIS Characterization Support Team maintains a number of good resources, including the MCST FAQ and MCST Calibration Guide.

Q: How do I determine which MODIS atmospheric product to order?

A: You first must decide whether global L3 summaries at 1 degree resolution will meet your needs. If so, you will find all MODIS Atmosphere parameters together in single daily (MOD08_D3/MYD08_D3), 8-day (MOD08_E3/MYD08_E3) or monthly (MOD08_M3/MYD08_M3) product files. However, you do have the option of ordering via LAADS Web a subset of one of these files containing only the parameters you select. If you decide to order higher resolution Atmospheric L2 products, please consult the Atmospheric Products Guide for more detailed product information.

Q: MODIS product file names seem a bit arcane. Is there a pattern here?

A: MODIS product file names contain the product ID, a date/time ID (usually the beginning of data acquisition), a collection ID and finally a processing date/time stamp. A quick reference guide can be found in the Guide to Atmospheric Product File Names.

Q: What is the foramt of the Distribution Notice (DN) files generated by LAADS?

LAADS can generate Distribution Notice (DN) files for subscriptions or LAADS Web orders using the FTP-Push delivery method. The DN files use the standard SIPS Distribution Notice format that is defined in "Interface Control Document between the EOSDIS Core System (ECS) and the Science Investigator-Led Processing Systems (SIPS), Volume 0 Interface Mechanisms". A standard DN file will list the destination FTP host and directory information, the order ID (REQUESTID) in LAADS, the time at which the order was pushed, and file name, file size, cksum type and cksum value for each product file in the order.

++++++++++

ORDERID: NONE
REQUESTID: 12588352
USERSTRING:
FINISHED: 06/13/2006 09:15:32

MEDIATYPE: FtpPush
FTPHOST: mtvs2
FTPDIR: /MYDATML2
MEDIA 1 of 1
MEDIAID:

        GRANULE:
        ESDT: MYDATML2

                FILENAME: MYDATML2.A2006158.2315.005.2006164085603.hdf
                FILESIZE: 3295302
                FILECKSUMTYPE: CKSUM
                FILECKSUMVALUE: 2538623922

        GRANULE:
        ESDT: MYDATML2

                FILENAME: MYDATML2.A2006158.1825.005.2006164085619.hdf
                FILESIZE: 2278554
                FILECKSUMTYPE: CKSUM
                FILECKSUMVALUE: 3462502739

Q: I see references to MODIS, Aqua and Terra. Please elaborate.

A: The Moderate Resolution Imaging Spectroradiometer, MODIS, instrument flies on two polar orbiting satellites. The first, Terra, has provided Land, Atmosphere and Ocean data products since late February 2000. The second, Aqua, started delivery science data in late June 2002.

Q: How do I choose between Aqua and Terra?

A: Generally, the same list of retrieved geophysical parameters is available from both missions. However, orbits and instrument calibration differences will often dictate which MODIS data source is most appropriate. The Univ. of Wisconsin SSEC maintains an excellent resource for both Terra Orbit Tracks and Aqua Orbit Tracks. These sites can be used to determine how closely a daily overpass samples any geographical point and to locate the 5 minute data granule that contains overpass data of interest. The MCST calibration team also maintains an Overflight Planner.

Q: What is the format of the product data files?

A: The standard format for MODIS products is HDF-EOS. This format is intended to be self-documenting, that is, with "metadata" imbedded within each product file. The bad news for anyone encountering HDF for the first time is that you are likely to be at an impasse until you acquire a special set of tools to break open and investigate these files. However, there is a lot of good news as well. There are a lot of mature tools that work well. They are provided free of charge and generally run on a range of computer platforms (Windows, Linux, OS X, Irix, etc.). LAADS Web also provides data in non-native formats such as GeoTIFF. HDF-EOS allows us to deliver geographical cut-outs, sub-sampled products, mosaic-ed products, etc, all in a format that is usable by these tools.

Q: How are product files compressed?

A: Product files as a whole are generally not compressed; but individual SDSs within each product file may be compressed, usually by using the hrepack utility. The same routine can be used to unchunk and decompress the SDSs, by choosing the compression method "NONE" and chunking info of "NONE". If $HDFBIN/hrepack doesn't exist on your system, go to the HDF Tools website to get a copy.

MODAPS uses a lossless compression method (GZIP) for compressing the SDSs. Therefore, compressing an HDF SDS will have no effect on the contents of the SDS as read by HDF library routines. Those routines check whether or not the SDS is compressed, and decompress the data, if necessary, without requiring any special action on the part of the user. A program that reads the SDS, and never checks whether it's compressed, or tries to write to it, will work the same with compressed data as with uncompressed ones.

Reading a compressed SDS is slower than reading an uncompressed SDS, but that effect can be minimized by what HDF calls "chunking", and we do chunk many of the larger SDSs that are compressed. Most programs that read a small number of scans at a time, or a large fraction of a scan each time, will not need any re-writing to achieve reasonable performance with chunked data. However, for the best performance a program can call the HDF library routine SDgetchunkinfo() in C, or sfgichnk() in Fortran, to determine the chunking parameters. Then it can call SDreadchunk() in C, or sfrchnk() in Fortran, to read (and decompress, if necessary) an entire chunk at a time.

However, any program that attempts to modify part of a compressed SDS may fail, without producing any error indications. This is because the HDF library only permits modification of the entire compressed SDS in a single write. Exception: if it calls SDchunkwrite() in C or sfwchnk() in Fortran, it can update one chunk at a time, rather than the entire SDS.

Q: What tools to do you recommend to crack open these HDF-EOS files?

A: Two tools are particularly indispensible for working with L1 or Atmosphere products. The first is "ncdump", available from the NCSA HDF Tools page. The command "ncdump -h file_name" displays all file metadata. Otherwise, "ncdump" will show you more than you want to see. The second recommended tool is HDFLook, which was developed by the Lille team with EOS project support. HDFLook is an interactive visualization tool that can also be run on the (Unix) command line to produce images or reformat HDF file data. (If you have to have binary files, HDFLook can, for example, create binary files containing single-parameter arrays.)

Q: Where do I find more useful tools?

A: A good starting point for L1 and Atmosphere products is the list of Atmosphere Discipline Recommended Tools. The GES-DISC also maintains a list of GES-DISC Recommended Tools. These lists contain a range of tools of vary complexity. Some require licensed software such as IDL. Please return to these sites as they are updated.

Q: Where do I go for pictures?

A: Each discipline maintains and image library. L1 and Atmospheric images can be found at the MODIS-ATMOS Image Gallery. There is a library of L3 daily global browse as well as RGB images of each granule using Bands 1,4,3 of the 1km Radiance product.

Q: Is there a page for up-to-date instrument status and information?

A: The EOS project maintains a Terra Events Page. Another resource is MODIS News and Current Status. For data production status, please check the LAADS Web Data Availability Chart.

Q: Please point me to more Atmospheric product information.

A: The best starting point is the MODIS Atmosphere Discipline Site. There you will find the Atmospheric Products Guide. The Product Change Summary for Collection 5 was prepared as a guide to product changes between collections 4 and 5. The team records the complete production history of each product in the Detailed Atmosphere Product Availability Calendar. If you want the complete story on how QA bits are used within each product, see QA Product Information for Collection 5.

Q: Where can I find details about the MODIS retrieval algorithms for Atmospheric products?

A: The MODIS Atmosphere Discipline Site is again the place to start. Two documents of particular interest are the Known Problems of Atmosphere Products and Validation of Atmosphere Products. You should also be aware of the Atmosphere Team Publications and Presentations. The GES-DISC also maintains a general resource known as the GES-DISC Parameter Information Pages.

Other more detailed algorithm references: Historical ATBDs (somewhat outdated) PGE03 Algorithm Change History PGE04 Algorithm Change History PGE05 Algorithm Change History PGE56 Algorithm Change History PGE57 Algorithm Change History PGE70 Algorithm Change History PGE83 Algorithm Change History