Validation Status

AQUA MODIS SNOW AND SEA ICE DATA PRODUCTS VALIDATION AND PRODUCT QUALITY STATUS UPDATE
24 February 2004

Collection 4 processing of MODIS Aqua products began on 2 February 2004. The Aqua snow and sea ice products differ from the Terra products because MODIS band 7 (2.1 µm) is used in place of MODIS band 6 (1.6 µm) in the Aqua products. Bands were changed because the many non-functional detectors in MODIS Aqua band 6 significantly decreased quality of the data. Snow reflectance is similar in bands 6 and 7 and a very similar snow or ice map can be obtained with band 7, however there are differences that are being analyzed as Collection 4 grows. Aqua and Terra validation and quality status are tracked and reported separately.

Aqua Collection 4 is being processed in forward and reprocessing streams. Forward processing began on 2 February 2004 (2004 033). Reprocessing began starting with Aqua data from 4 July 2002 (2002 185). Aqua Collection 4 data products were released to users on 13 February 2004 by NSIDC.

There are periods; minutes, hours or days, in the Aqua data record missing due to various reasons, e.g. sensor, platform or production. Data outages, for Aqua and Terra, have been documented and are reported by the Goddard Earth Sciences (GES) DAAC.

Aqua Collection 4 Snow Products
MYD10_L2   -   Validation status: Provisional.
Aqua MODIS band 7 is used in the algorithm. The test for snow in dense vegetation in the algorithm was disabled because it was observed to result in frequent erroneous snow mapping in some situations. Disabling that test reduced but, has not entirely eliminated snow errors. Notable instances of error have been observed associated with cloud edges over vegetated surfaces. Investigation is continuing. The cloud mask product, MYD35_L2, used as input to the MYD10_L2 algorithm also changed to use of band 7. The effect of that change relative to snow/cloud discrimination is being investigated.

MYD10A1   -   Validation status: Provisional.
Daily snow maps have improved quality due to a new scoring selection algorithm based on the solar elevation, distance from nadir and coverage of an observation in a grid cell of the projection. Reduction of erroneous snow mapping in high northern latitudes during summer months has occurred. Erroneous snow mapping associated with cloud edges over vegetated surfaces in MYD10_L2 is carried forward into the MYD10A1 product. A snow albedo data array is contained in the product. Snow albedo is calculated for all snow observations in the corresponding snow map. Snow albedo has a beta validation status.

MYD10C1   -   Validation status: Provisional.
Daily global snow maps have generally improved quality. Improvements and errors from MYD10A1 are carried forward into the MYD10C1 product.

MYD10A2   -   Validation status: Provisional.
Eight-day snow maps have quality similar to the MYD10A1 products. The algorithm generates a maximum snow extent map so erroneous snow in MYD10A1 is maximized in the eight-day snow map.

MYD10C2   -   Validation status: Provisional.
Eight-day global snow maps show improved continuity in features mapped. Improvements from MYD10A2 are apparent in this product, as are the errors also. Erroneous mapping of low snow fractions decreases quality of the eight-day snow map.

Aqua Collection 4 Sea Ice Products
MYD29   -   Validation status: Provisional.
Aqua MODIS band 7 is used in the algorithm. Good quality has been observed in the sea ice maps; however, investigation of effects of the switch to band 7 is continuing. The cloud mask product, MYD35_L2, used as input to the MYD29 algorithm also changed to use of band 7. The effect of that change relative to sea ice/cloud discrimination is being investigated. The ice surface temperature (IST) was not affected by the switch to band 7 except, possibly indirectly by the cloud mask switch to band 7. Validation status is set at provisional until further validation work specific to Aqua IST maps can be completed.

MYD29P1D   -   Validation status: Provisional.
Daily sea ice maps have greatly improved quality due to the good quality of the MYD29 V004 products.

MYD29E1D   -   Validation status: Provisional.
Data quality greatly improved because of improvements in the MYD29 V004 and MYD29P1D V004products. In polar regions at the boundary of polar darkness and daylight, inconsistency in features mapped is observed. That inconsistency appears to be related to how observations in darkness are processed in the input data products.

TERRA MODIS SNOW DATA PRODUCTS COLLECTION 4 AND COLLECTION 3 VALIDATION AND PRODUCT QUALITY STATUS
28 January 2003

All the MODIS Collection 3 and Collection 4 snow products are considered to be Validated, Stage 2. Collection 4 is given the same validation status as Collection 3 because only minor revisions that affect scientific quality of the products were made in the algorithms for Collection 4. Analysis of the four science test runs made for Collection 4 processing revealed that the approved versions of the algorithms gave results consistent with Collection 3 performance so were given the same validation status (Table 1). A caveat to validation of Collection 4 is, the period from first day of data acquisition (24 February 2000) to the start of Collection 3 (Table 2) is first time public release of the snow data products and although the MODIS Level 1B data is declared validated, it may have characteristics during that time period that affect the quality of the snow data products. Analysis of the snow data products is an ongoing activity. Specific information on the science quality of the snow data products is reported in the “ScienceQualityFlagExplanation”. The URL for the quality assessment site is given in the product metadata and is linked to from the EDG when ordering data. The “ScienceQualityFlagExplanation” is changed in response to analysis results and should be checked for updated information.

Data dates of Collection 3 and Collection 4 and validation status are presented in Table 1. The forward and reprocessing data production streams of Collection 4 are shown separately to show the time periods in each. Collection 4 begins on 24 Feb 2000 and continues to the present. Collection 3 begins on different dates for the snow data products. In each product, the differences between those start dates are periods for which the data will be publicly available for the first time (Table 2).

TERRA MODIS SEA ICE DATA PRODUCTS COLLECTION 4 AND COLLECTION 3 VALIDATION STATUS
12 February 2003

All the MODIS Collection 3 and Collection 4 sea ice products are considered to be Validated, Stage 1. Collection 4 is given the same validation status as Collection 3 because analysis of data from the four science test runs made for Collection 4 processing revealed that the revised versions of the algorithms gave results consistent with or better than Collection 3 performance (Table 1). A notable improvement in the quality of the daily sea ice product, MOD29P1D, has been observed due to inclusion of a new scoring algorithm that results in coherent mapping of sea ice. During the course of Collection 3 processing, changes were made in the L2G mapping algorithm and product and in the sea ice algorithm that resulted in a loss of coherence in the sea ice products for long time periods in Collection 3. A caveat to validation of Collection 4 is, the period from first day of data acquisition (24 February 2000) to the start of Collection 3 (Table 2) is first time public release of the sea ice data products and only limited investigation of product quality for that time period has been accomplished. Although the MODIS Level 1B data is declared validated, it may have characteristics prior to November 2000 that affect the quality of the sea ice data products.

Analysis of the quality of the sea ice data products is an ongoing activity. Specific information on the science quality of the sea ice data products is reported in the “ScienceQualityFlagExplanation”. The URL for the quality assessment site is given in the product metadata and is linked to from the EOS Data Gateway (EDG) when ordering data. The “ScienceQualityFlagExplanation” is changed in response to analysis and should be checked for updated information. In the MOD29 and MOD29P1D data products there are two instances of the “ScienceQualityFlagExplanation”, one for sea ice determined by reflectance data and one for IST written in the metadata. Information on both is posted at that URL.

Data dates of Collection 3 and Collection 4 and validation status are presented in Table 1. The forward and reprocessing data production streams of Collection 4 are shown separately to show the time periods in each. Collection 4 begins on 24 February 2000 and continues to the present. Collection 3 begins on different dates for the sea ice data products. In each product, the differences between those start dates are periods for which the data will be publicly available for the first time (Table 2).

Publications dealing with Validation

Hall, D.K., G.A. Riggs, V.V. Salomonson, N. DiGiromamo and K.J. Bayr, 2002: "MODIS Snow-Cover Products," Remote Sensing of Environment, 83:181-194.

Hall, D.K., R.E.J. Kelly, G.A. Riggs, A.T.C. Chang and J.L. Foster, 2002: "Assessment of the relative accuracy of hemispheric-scale snow-cover maps," Annals of Glaciology, V.34:24-30

Hall, D.K., R.E.J. Kelly, A.T.C. Chang, J.L. Foster and J.Y.L. Chien, 2002: "Analysis of Relative Errors in Snow Maps in North America, Winter 2001-02," (abstract only), presented at Progress in Electromagnetics Research Symposium 2002 (PIERS 2002), Cambridge, MA, 1 July 2002.

Riggs, G.A. and D. K. Hall, in press: "Reduction of Cloud Obscuration in the MODIS Snow Data Product," Presented at the 59th Eastern Snow Conference, 5-7 June 2002, Stowe, VT.

Hall, D.K., R. Solberg and G.A. Riggs, 2002: "Validation of satellite snow cover maps in North America and Norway," Proceedings of the 59th Eastern Snow Conference, 5-7 June 2002, Stowe, VT.

Data Product Validation and Quality Assessment Links