Home

MODIS Snow Products User Guide to Collection 5 George A. Riggs

1. These latitude and longitude pairs of points when connected in a clockwise series form a polygon of the swath coverage Always represents the outer ring of coverage Beginning and ending times of the first and last scan line in the swath Formats are yyyy mm dd Lr 0000 sss RangeEndingTime 20 30 00 000000 00000000 30 00 000000 Version of production PGEVersion 5 0 6 generation executable PGE AssociatedSensorShortName MODI 000 Sensor name AssociatedPlatformShortName Terra Platform name Instrument and sensor AssociatedInstrumentShortName MODIS name are the same Product Specific Attributes PSA QAPERCENTGOODQUALITY 100 Summary quality of data range checks done in the QAPERCENTOTHERQUALITY 0 algorithm HORIZONTALTILENUMBER 09 eo VERTICALTILENUMBER 04 In latitude direction 0 17 31 80 TilelD SNOWCOVERPERCENT 51009004 06 Format is pshhhvvv p projection code S hhh horizontal tile size 1 is full size number vvv vertical tile number Summary percentage of snow covered land The ArchiveMetadata 0 global attribute contains information relevant to version of the algorithm production environment and geographic location of the data product Contents are described in Table 19 Table 19 Listing of objects in ArchiveMetadata 0 the global attribute in MOD10A1 Object Name CHARACTERISTICBINA
2. 0 Monthly Global Snow Cover 5 MOD10CM 71 of 80 Comment Filename of product Format is EDST Ayyyyddd vw yyyydddhhmmss hdf Ayyyyddd hhmm acquisition date and time in UTC vvv collection version yyyydddhhmmss date and time of production hdf HDF file extension Date and time the file was produced Format is yyyy mm ddThh mm ss sssZ Day means all data in daylight Both means that daylight and darkness were included Reprocessed means data has been processed before Processed once means this is the first processing of the data Version of algorithm delivered from the SCF Expect that the product will be reprocessed again with an improved algorithm This is meaningless information Original plan was for this metadata to be set updated by investigator after evaluation validation however that plan was dropped and this metadata is not set updated See ScienceQualityFlagExplanation for current information No automated QA checks made during execution of the algorithm Default setting because no automated QA checks are done URL where updated information on science QA should be posted Amount of data missing from the swath Amount of land in the swath obscured by clouds QA parameters given apply to the snow cover data Indicates the EOSDIS Collection ESDT name of product MOD10C1 A2005244 004 2005247012647 hdf MOD10C1 A2005246 004 2005249
3. MODIS Terra Snow Cover Daily Descriptive name of the product May be LongName L3 Global 0 05Deg CMG displayed as the product name in the EOS Data Gateway or other dataset search tools Moderate Resolution Imaging InstrumentName SpectroRadiometer Long name of MODIS PLATFORMSHORTNAME Terra GLOBALGRIDCOLUMNS 7200 GLOBALGRIDROWS 3600 Processing Center MODAPS MODIS Adaptive Processing System ProcessingDateTime 2006 02 21723 54 38 0000007 Gate 0 Processing Format is yyyymm SPSOParameters none Archaic and meaningless DESCRRevision 5 0 Descriptor file associated with the PGE Processing Environment IRIX64 mtvs3 6 5 10070055 IP35 7 DESCRRevision 5 2 Descriptor file associated with the PGE The StructMetadata 0 global attribute is created by the HDF EOS toolkit to specify the mapping relationships between the map projection and the snow cover data SDSs Mapping relationships are unique in HDF EOS and are stored in the product using HDF structures Description of the mapping relationships is not given here Use of HDF EOS toolkit other EOSDIS supplied toolkits DAAC tools or other software packages may be used to geolocate the data or to transform it to other projections Map projection parameters are from the GCTP Listing of the global attribute StructMetadata 0 in MOD10A1 StructMetadata 0 StructMetadata O GROUP SwathStructure END GROUP SwathStructure GROUP GridStructure GROUP GRID 1 GridN
4. Snow Evaluation and Errors The daily snow product snow cover map and fractional snow cover inherits snow errors associated with the observation selected from the MOD10 L2 swath product In this version of the algorithm no attempt was made to screen or correct snow errors in the input data Efforts were focused on reducing the snow errors in the MOD10 L2 algorithm which would then result in reduction of snow errors in the MOD10A1 product That approach has resulted in a reduction of snow errors being passed into the MOD10A1 snow cover map The mapping of the pixel observations from MOD10 L2 into the grid cells in the L2G process may result in a pixel being mapped into more that one grid cell If that is the situation with an erroneous snow observation then it is possible that a single erroneous snow observation will be mapped into and selected for one or more cells in the MOD10A1 snow map In that situation the extent of erroneous snow is seen to increase These snow errors are problematic to users 28 of 80 being readily apparent in some regions and seasons but not in others Apparent errors may be screened by users by use of screens of their own design A prominent feature along coast lines in some areas e g Arctic regions during the summer season is a coating of snow The snowy coastline is a result of swath image and land water mask misalignment originating in the MOD10 L2 product Until the misalignment situation is resolved these error
5. 5 55 Sensor name AssociatedPlatformShortName Terra Platform name Instrument and sensor AssociatedlnstrumentShortName MODIS name are ithe same Product Specific Attributes PSA QAPERCENTGOODQUALITY 100 Summary quality of data range checks done in the QAPERCENTOTHERQUALITY 0 algorithm Summary percentage of SNOWCOVERPERCENT 11 snow covered land Table 42 Listing of objects in ArchiveMetadata 0 the global attribute in MOD10C2 Object Name Typical Value Comment AlgorithmPackageAcceptanceDate 01 2005 AlgorithmPackageMaturityCode Normal Algorithm version information Format is i i mm yyyy AlgorithmPackageName MOD PR10A1 AlgorithmPackageVersion 5 MODIS Terra Snow Cover 8 Day L3 Global 0 05Deg CMG Descriptive name of the product May be displayed as the product name in the EOS Data Gateway or other dataset search tools InstrumentName 2 Imaging Long name of MODIS PLATFORMSHORTNAME Terra GLOBALGRIDCOLUMNS 7200 GLOBALGRIDROWS 3600 Processing Center MODAPS MODIS Adaptive Processing System 2005 03 13T07 30 05 000000Z Date of processing Format is yyyy mm ProcessingDateTime dd Thhimmss essz SPSOParameters none Archaic and meaningless DESCRRevision 5 0 Descriptor file associated with the PGE 3 T Processing done in either UNIX or Linux Processing Environment IRIX64 mtvs3 6 5 10070055 IP35 environment DESCRRevision 5 2 Descriptor ass
6. fractional snow cover in the L2G product using the same observation scoring algorithm as used for the daily snow cover map The fractional snow map for the day is stored in the Fractional Snow Cover SDS Snow albedo is calculated for the visible and near infra red bands using the MODIS land surface reflectance product as input Table 11 lists the inputs for the snow albedo algorithm An anisotropic response function is used to correct for anisotropic scattering effects of snow in non forested areas Snow covered forests are assumed to be Lambertian reflectors Land cover type is read from the MODIS land cover product Slope and aspect data for the correction is derived from the Global 30 Arcsecond GTOPO30 digital elevation model DEM are stored for each tile as ancillary data files The narrow band albedos are then converted to a broadband albedo for snow Description of the snow albedo algorithm is given in Klein and Stroeve 2002 Snow albedo is calculated only for the cells that correspond to snow cover in the Snow Cover Day Tile Snow albedo is stored in the Snow Albedo Daily Tile SDS Table 11 MODIS data product inputs to the MODIS daily snow algorithm ESDT Long Name Data Used MODIS Terra Snow Cover Daily L2G Snow cover fractional MOD10L2G Global 500m SIN Grid 2 snow spatial MODIS Terra Geolocation Angles Solar and sensor MODMGGAD Daily L2G Global 1km SIN Grid Day geometry Number of MODIS Terra Observation Pointers obse
7. unpublished evaluations That estimate is based on best conditions for the algorithm however in conditions difficult to calculate snow albedo e g steep mountain terrain the snow albedo error is likely to be very large Updates to snow albedo evaluation and validation will be posted on the snow project website Global Attributes There are 11 global attributes in the MOD10A1 product three are ECS defined CoreMetadata 0 ArchiveMetadata 0 and StructMetadata 0 and the others are specific to the product These global attributes serve different purposes such as search and order of products mapping and product version tracking and evaluating a product The ECS defined attributes are written as very long character strings in parameter value language PVL format Descriptions of the global attributes are given in the following tables CoreMetadata 0 and ArchiveMetadata 0 are global attributes in which information compiled about the product during product generation is archived StructMetadata 0 contains information about the swath or grid mapping relevant 29 of 80 to the product A user wanting detailed explanations of the global attributes and related information should query the EOSDIS related web sites Table 18 Listing of objects in the global attribute CoreMetadata 0 in MOD10A1 Object Name LocalGranulelD ProductionDateTime DayNightFlag ReprocessingActual LocalVersionID ReprocessingPlanned ScienceQualityFlag
8. 2006043004036 hdf MOD10A1 A2003200 h23v15 005 2006043000237 hdf MOD10A1 A2003200 h24v15 005 2006043001 423 hdf EASTBOUNDINGCOORDINATE 180 0 WESTBOUNDINGCOORDINATE 180 0 Mid D dd SOUTHBOUNDINGCOORDINATE 90 0 NORTHBOUNDINGCOORDINATE 90 0 ZONEIDENTIFIER Other Grid System LOCALITYVALUE Global RangeEndingDate 2003 07 19 RangeEndingTime 23 59 59 2003 07 19 Beginning and ending times the day Formats RangeBeginningTime 00 00 00 000 00 00 yyyy mm dd hh mm ss i Names of MODIS data input InputPointer files Version of production generation executable w o AssociatedSensorShortName AssociatedPlatformShortName Terra Platform name Instrument and sensor AssociatedInstrumentShortName MODIS namere ihe sam Product Specific Attributes PSA QAPERCENTGOODQUALITY 100 Summary quality of data range checks done in the QAPERCENTOTHERQUALITY 0 algorithm Summary percentage of SNOWCOVERPERCENT 31 snow covered land PGEVersion 5 0 5 i 45 of 80 Table 28 Listing of objects ArchiveMetadata 0 the global attribute in MOD 10C1 Object Name Typical Value Comment AlgorithmPackageAcceptanceDate 05 2006 AlgorithmPackageMaturityCode Normal i i Algorithm version information Format is mm yyyy AlgorithmPackageName MOD_PR10A1 AlgorithmPackageVersion
9. 9999999964079 104 421704737634 140 015144391787 Date of processing Format is yyyy mm ddThh mm ss sssZ Archaic and meaningless Eastern western northern and southern most points of the swath Format is decimal degrees Processing done in either UNIX or Linux environment Linux minion5024 2 6 8 1 24mdksmp 1 SMP Thu Jan 13 Processing Environment 23 11 43 MST 2005 i686 Descriptor file DESCRRevision 5 2 associated with the PGE The StructMetadata 0 global attribute is created by the HDF EOS toolkit to specify the mapping relationships between the map projection and the snow cover data SDSs Mapping relationships are unique in HDF EOS and are stored in the product using HDF structures Description of the mapping relationships is not given here Use of HDF EOS toolkit other EOSDIS supplied toolkits DAAC tools or other software packages may be used to geolocate the data or to transform it to other projections Map projection parameters are from the GCTP Listing of the global attribute StructMetadata 0 in MOD10A1 StructMetadata 0 GROUP SwathStructure END GROUP SwathStructure GROUP GridStructure GROUP GRID 1 GridName MOD Grid Snow 500m XDim 2400 YDim 2400 UpperLeftPointMtrs 10007554 677000 5559752 598333 LowerRightMtrs 8895604 157333 4447802 078667 Projection GCTP_SNSOID ProjParams 6371007 181000 0 0 0 0 0 0 0 0 0 0 0 0 SphereCode 1 GridOriginZHDFE GD UL GROUP Dimension END GROUP Dimension GRO
10. AutomaticQualityFlagExplanation AutomaticQualityFlag ScienceQualityFlagExplanation Sample Value MOD10A1 A2003201 h09v04 005 2006043034028 hdf 2006 02 12T03 41 45 000Z Day reprocessed SCF V5 0 5 further update is anticipated Not investigated No automatic quality assessment done in the PGE Passed See http landweb nascom nasa gov cgi bin QA WWW qaFlagPage cgi sat terra the product Science Quality status 30 of 80 Comment Filename of product Format is EDST Ayyyyddd hnnvnn v vv yyyydddhhmmss hdf Ayyyyddd hhmm acquisition date and time in UTC hnnvnn horizontal and vertical tile number vvv collection version yyyydddhhmmss date and time of production hdf HDF file extension Date and time the file was produced Format is yyyy mm ddThh mm ss sssZ Day means entire swath in daylight Both means that part of swath lies in darkness Reprocessed means data has been processed before Processed once means this is the first processing of the data Version of algorithm delivered from the SCF Expect that the product will be reprocessed again with an improved algorithm This is meaningless information Original plan was for this metadata to be set updated by investigator after evaluation validation however that plan was dropped and this metadata is not set updated See ScienceQualityFlagExplan ation for current information No automated QA
11. Passed QA checks are done See http landweb nascom nasa gov cgi bin QA WWW qaFlagPage cgi sat terra for the product Science Quality status 0 ane of L1B data missing from the 000000000002 7 land the swath obscured Parameternamn Sowo 00000000002 given apply to the snow EqutorCrossngDate pooto 2 parameter Format yyyy mm OrbitNumber 8335 Orbital parameter EquatorCrossingLongitude 106 330685 1 parameter Decimal degrees VersionID 5 Indicates the EOSDIS Collection ShortName MOD10_L2 ESDT name of product MODO2HKM A2003198 1945 005 20060360528 19 hdf MOD021KM A2003198 1945 005 200603605281 ScienceQualityFlag Not being investigated URL where updated information on science QA should be posted ScienceQualityFlagExplanation InputPointer Names of MODIS data input files 18 of 80 9 hdf MOD35 L2 A2003198 1945 005 20060360701 1 1 hdf MOD03 A2003198 1945 005 20060351 12242 hd f These latitude and longitude pairs of points when connected in a clockwise series form a polygon of the swath coverage Always represents the outer ring of coverage AREE DNE DNRA 148 p GringPointSequenceNo 1 2 3 4 ExclusionGRingFlag N RangeBeginningDate 2003 07 17 RangeBeginningTime 19 45 00 000000 RangeEndingDate 2003 07 17 RangeEndingTime 19
12. Table 45 Local attributes for Snow Cover Monthly CMG Attribute name Definition Value Monthly snow long name Long Name of the SDS cover extent 5km units SI units of the data if any none 68 of 80 format coordsys valid range _FillValue Mask_value Night_value Cell resolution Antarctica_sno w_note Key How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans For seasonal darkness Nominal grid cell resolution Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names Quality Assessment Minimal QA is applied to the data during processing By default the thematic QA is set to good quality and is changed only if all the input data is bad or if a masked class e g ocean is applied Snow Spatial QA Minimal QA for each cell of the grid is written in this SDS Table 46 Local attributes for Snow Spatial QA Attribute name long name units Definition Long Name of the SDS SI units of the data if any 69 of 80 I3 latitude longitude 0 100 255 254 211 0 05 deg Antarctica deliberately mapped as snow 0 100 of snow in cell 211 night 250 cloud 253 no decision 254 water mask 255 fill Value Thematic QA map of the monthly Snow none How th
13. Water mask la nd threshold 76 Antarctica confi dence index n ote Key Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans Nominal grid cell resolution Decision point to process a cell as land or water Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names Snow Evaluation and Errors An indicator of quality of the MOD10A1 observations that were mapped into a CMG cell is reported in the Snow Spatial QA SDS This indicator is a summary representative of the quality of the MOD10A1 observations that were mapped into the CMG cell latitude longitude 0 100 255 254 0 05 deg 12 00000 Antarctica deliberately mapped as snow Confidence index set to 100 0 100 confidence index value 107 lake ice 111 night 250 cloud obscured water 253 data not mapped 254 water mask 255 fill Table 26 Local attributes for Snow Spatial QA SDS Attribute name long name units Definition Long Name of the SDS SI units of the data if any 41 of 80 Value Snow cover per cell QA none format coordsys valid range _ FillValue Mask value Cell resolution Water mask la nd threshold 76 Antarctica QA _ note Key How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values
14. checks made during execution of the algorithm Default setting because no automated QA checks are done URL where updated information on science QA should be posted 67 Amount of L1B data QAPercentMissingData 0 missing from the swath Amount of land in the QAPercentCloudCover 18 swath obscured by clouds QA parameters given apply to the snow cover data i ParameterName Snow Cover Daily Tile 5now Albedo Daily Tile EquatorCrossingDate 2003 07 EquatorCrossingTime 17 21 47 571376 OrbitNumber 19082 EquatorCrossingLongitude 103 091848200135 Indicates the EOSDIS Version D 5 Collection ShortName MOD10A1 ESDT name of product MOD10L2G A2003201 h09v04 005 2006043032816 hdf MODMGGAD A2003201 h09v04 005 2006043030423 hdf MODPTHKM A2003201 h09v04 005 2006043030339 hdf MODO9GHK A2003201 h09v04 005 2006043031 930 hdf MOD1201 A2001001 h09v04 004 2004358134052 hdf m 117 746445975456 140 795234672207 124 615349244084 104 235445821904 nw 39 7342308150748 49 9394187999602 firigeoiniFautude 50 1159178280076 39 8623890159424 GringPointSequenceNo 1 2 3 4 ExclusionGRingFlag N RangeBeginningDate 2003 07 20 RangeBeginningTime 17 10 00 000000 22 71 2003 07 20 Orbital parameters Format yyyy mm dd Format hh mm ss ssssss Decimal degrees format Data given for each swath input Names of MODIS data InputPointer input files
15. cover is found for any day in the period then the cell in the 49 of 80 Maximum Snow Extent SDS is labeled as snow If no snow is found but there is one value that occurs more than once that value is placed in the cell e g water on five days cloud on one land on one and night on one would be labeled water Otherwise if mixed observations occur e g land and cloud over multiple days the algorithm is biased to clear views in the period and will label a cell with what was observable The logic minimizes cloud cover extent in that a cell would need to be cloud obscured for all days of observation to be labeled as cloud If all the observations for a cell are analyzed but a result is not reached then that cell is labeled as no decision A chronology of snow occurrence is recorded in the Eight Day Snow Cover SDS On days that snow is found the bit corresponding to that day eight days across the byte from right to left is set to on The input days are ordered from first to last day including placing any missing days in the order Table 31 MODIS data product inputs to the MOD10A2 snow algorithm ESDT Long Name Data Used MODIS Terra Snow Cover Daily L3 Global 500m SIN Grid Snow cover MOD10A1 The algorithm will generate a product if there are two or more days of input available If there is only a single day of input the eight day period the product will not be produced All eight days of input may sometimes not be available due
16. filter are assigned 096 snow for the month Cells with a low magnitude are considered suspect of being erroneous snow originating in the MOD10 L2 algorithm and being propagated through the sequence of snow products The magnitude of snow is calculated as an average snow for all days with snow passing the first filter of Cl gt 70 For example cell has 20 days with Cl 100 10 days have 100 snow and 10 days have 0 snow the mean monthly snow 10 100 10 0 20 50 The second filter would be calculated as days of snow Cl days of snow 10 100 10 100 That average is retained because the average snow magnitude was 10 Cell B also has 20 days with Cl 100 however the 10 days of snow are all 5 In this case the snow magnitude is 5 10 10 5 thus the cell is filtered out and the monthly snow average is set to 0 Minimal QA is applied to the data By default the QA is set to good quality and is changed only if all the input data is bad or if a masked class e g ocean is applied Table 44 MODIS data product inputs to the MOD10CM snow algorithm ESDT Long Name Data Used MODIS Terra Snow Cover Daily L3 Snow cover cloud cover MOD10C1 0 05Deg CMG Cl Scientific Data Sets Snow_Cover_Monthly_CMG The mean monthly fractional snow cover data is stored in this SDS Mean monthly fractional snow is reported in the range 0 100 Fig 10 Other features are mapped with specific values e g water feature 254
17. format coordsys valid range FillValue Mask value Not processed value Night value Water mask la nd threshold 76 Antarctica sno w note Key SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans For seasonal darkness Decision point to process a cell as land or water Antarctica masked as perennial Snow cover Key to meaning of data in the SDS HDF predefined attribute names 61 of 80 percentage for the eight day snow map none I3 latitude longitude 0 100 255 254 252 111 12 00000 Antarctica deliberately mapped as snow Cloud value set to 252 0 100 of cloud in cell 107 lake 111 night 250 cloud obscured water 253 data not mapped 254 water mask 255 fill Quality Assessment Snow Spatial QA The data is indicative of the overall quality of data in the cell In Collection 5 the QA is not fully utilized The QA value is set to good quality by default and is not changed unless the input data are unusable data The logic for determining setting the QA of the eight day product is being discussed Table 40 Local attributes for Snow Spatial QA SDS Attribute name long name units format coordsys valid range FillValue Mask val
18. from the multiple observations mapped to a cell of the MOD10 L2G gridded product from the MOD10 L2 swath product In addition to the snow data arrays mapped in from the MOD10 L2G snow albedo is calculated There are four SDSs or data fields of snow data snow cover map fractional snow cover snow albedo and QA in the data product file Algorithm Description The daily snow cover map is constructed by examining the many observations acquired for a day mapped to cells of the grid by the L2G algorithm A scoring algorithm is used to select an observation for the day The scoring algorithm is based on location of pixel and solar elevation Observations are scored based on distance from nadir area of coverage in a grid cell and solar elevation The object of the scoring is to select the observation closest to local noon time highest solar elevation angle nearest to nadir with greatest coverage that was mapped into the grid cell Form of the scoring algorithm is Score 0 5 solar elevation 0 3 distance from nadir 0 2 observation coverage Results of the snow cover algorithm a daily snow map of the region covered by the tile are stored in the Snow Cover Day Tile and per cell QA data for that snow map is stored in the Snow Spatial QA SDS The snow cover data are 23 of 80 stored as coded integer values with values being the same as assigned MOD 10 L2 Daily fractional snow cover is determined from the many observations of
19. is set to on Across a byte the days are ordered from right to left bit O corresponds to day 1 of the eight day period bit 1 corresponds to day 2 of the eight day period bit 7 corresponds to day 8 of the eight day period A bit setting of off could mean that data for that day was missing or that cloud was observed or that snow was not observed HDF predefined and custom local attributes are stored The HDF predefined attributes may be used by some software packages The custom local attributes are specific to the data in the SDS Local attributes are listed in Table 33 cartesian 0 254 255 0 2146587 2002 551 O missing data 1 no decision 11 night 25 no snow 37 lake 39 ocean 50 cloud 100 lake ice 200 snow 254 detector saturated 255 fill Table 33 Local Attributes for Eight Day Snow Cover SDS Attribute name long name Definition Long Name of the SDS 51 of 80 Value Eight day snow cover chronobyte units format coordsys valid range FillValue Key SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Key to meaning of data in the SDS HDF predefined attribute names Global Attributes bit I3 cartesian 0 255 0 Snow occurrence in chronological order Day in period ordered as 87654321 corresponds
20. of misalignment of the mask and imagery Snow mapping and lake ice mapping errors do occur in many situations These errors are a very low amount commonly in the lt 0 001 percentage range of total pixels processed in a swath Misidentification of rivers or lakes either mapped or un mapped in the land water mask as snow or lake ice may also occur if the water has high turbidity or if it is shallow with a bright bottom Those conditions may have an NDSI value in the snow range and have characteristics e g visible reflectance amount similar to snow that are not blocked by the screens in the algorithm Partially cloud obscured water bodies that are identified as probably clear can also sometimes be erroneously identified as ice covered for similar reasons The code was revised so that the screen for surface temperature is also applied to water body pixels identified as snow covered Application of that screen in those situations has decreased significantly the snow errors associated with water bodies especially during the warm seasons Low illumination snow errors Under low solar illumination conditions when an acquisition is hours away from the local solar noon e g during boreal summer or an acquisition is near or 15 of 80 includes the day night terminator snow errors can occur Algorithm processing takes the day night flag from the cloud mask MOD35 L2 which defines daylight as an observation with solar zenith angle 85 degrees Low
21. resolution of the Climate Modeling Grid CMG cells The eight day snow cover product MOD10A2 is an eight day composite of MOD10A1 to show maximum snow extent The global eight day snow cover product MOD10C2 is created by assembling MOD10A2 daily tiles and binning the 500 m cell observations to the 0 05 spatial resolution of the CMG The monthly snow cover product MOD10CM is a composite of the daily MOD10C1 maps for a month to map the maximum monthly snow cover Table 1 Summary of the MODIS snow data products Earth Science Data Type ESDT MOD10 L2 MOD10L2G MOD10A1 MOD10A2 MOD10C1 MOD10C2 MOD10CM Product Level L2 L2G L3 L3 L3 L3 L3 Nominal Data Array Dimensions 1354 km by 2000 km 1200km by 1200km 1200km by 1200km 1200km by 1200km 360 by 180 global 360 by 180 global 360 by 180 global Spatial Resolution 500m 500m 500m 500m 0 05 by 0 05 0 05 by 0 05 0 05 by 0 05 4 of 80 Temporal Resolution swath scene day of multiple coincident swaths day eight days day eight days month Map Projection None lat lon referenced Sinusoidal Sinusoidal Sinusoidal Geographic Geographic Geographic File Format of Snow Products The MODIS snow products are archived in Hierarchical Data Format Earth Observing System HDF EOS format files HDF developed by the National Center for Supe
22. shadowi ng 252 landmask mismatch 253 BRDF failure 254 non production mask HDF predefined attribute names Quality Assessment Spatial QA data corresponding to the snow cover observation selected for the daily snow cover map is also selected and mapped into the Snow Spatial QA SDS Table 17 Local attributes with Snow Spatial QA SDS Attribute name Value Spatial QA of the observation Definition long name Long Name of the SDS 27 of 80 units SI units of the data if any none How the data should be viewed Fortran format notation x Coordinate system to use for coordsys i e data cartesian x Max and min values within a vain Tangs selected data range ES FillValue Data used to fill gaps in the 255 swath 0 good quality 1 other quality 252 Antarctica Key Key to meaning of data in the mask 253 land SDS mask 254 mask saturated 255 fill HDF predefined attribute names Snow albedo specific QA is not reported in Collection 5 because ways of expressing the QA of the snow albedo result are being investigated Refer to the snow project website for validation information It is anticipated that future evaluation and validation of snow albedo will lead to the definition and setting of QA data Fractional snow specific QA data is also not reported because evaluation and validation of the product has not been completed refer to the snow project website for validation information
23. solar illumination conditions are processed for snow without consideration of the amount of radiation reaching the surface That was originally by design not limiting processing for certain conditions Analysis has revealed that low illumination of some surface features notably boreal vegetation types results in reflectance amounts and features that may be confused with that of snow Those same features under high illumination conditions near solar noon do not exhibit reflectance features similar to snow and are not mapped as snow The low amount of radiation on the surface and consequently lower reflection from the features can cause them to have an NDSI in the snow range so are erroneously identified as snow Because the low reflectance across the spectrum combined with the nature of a ratio can result in relatively small differences between the band 6 and band 4 to have a large NDSI ratio that may look like snow to the algorithm In the swath product erroneous snow mapping caused by low illumination conditions may contribute up to around 596 error based on count of land pixels analyzed for snow in a swath Erroneous snow caused by low illumination conditions was carried forward into the daily snow product MOD10A1 and consequently the MOD10C1 snow product in 004 prior to 13 September 2004 That decreased the quality of the MOD10A1 and MOD10C1 snow maps A reason that those erroneous snow observations were mapped into MOD10A1 was that the algor
24. tile number vvv collection version yyyydddhhmmss date and time of production hdf HDF file extension Date and time the file was produced Format is yyyy mm ddThh mm ss sssZ Day means entire swath in daylight Both means that part of swath lies in darkness Reprocessed means data has been processed before Processed once means this is the first processing of the data Version of algorithm delivered from the SCF Expect that the product will be reprocessed again with an improved algorithm This is meaningless information Original plan was for this metadata to be set updated by investigator after evaluation validation however that plan was dropped and this metadata is not set updated See ScienceQualityFlagExplanat ion for current information No automated QA checks made during execution of the algorithm Default setting because no automated QA checks are done URL where updated information on science QA should be posted 67 Amount of data missing QAPercentMissingData 0 from the swath Amount of land in the swath QAPercentCloudCover 26 obscured by clouds ParameterName Global Snow Cover QA parameters given apply to the snow cover data Indicates the EOSDIS VersionID 5 Collection ShortName MOD10C1 ESDT name of product MODAPSops3 PGE AM1 M coeff PGE67 MOD PR10C2 cmgTL5km global anc hdf MOD104A1 A2003200 h16v00 005 20060430038193 hdf MOD10A1 A2003200 h17v00 005
25. to bit order of 76543210 Bit value of 1 means snow was observed Bit value of 0 means snow was not observed ECS global attributes of CoreMetadata 0 ArchiveMetadata 0 and StructMetadata 0 are listed in Tables 34 and 35 and by listing Other global attributes are given in Table 36 Table 34 Listing of objects in the global attribute CoreMetadata 0 MOD10A2 Object Name LocalGranulelD ProductionDateTime Sample Value MOD10A2 A2003201 h11v05 005 2005071232605 hdf 2005 03 12T23 26 10 000Z 52 of 80 Comment Filename of product Format is EDST Ayyyyddd hnnvnn vvv y yyydddhhmmss hdf Ayyyyddd hhmm acquisition date and time in UTC hnnvnn horizontal and vertical tile number vvv collection version yyyydddhhmmss date and time of production hdf HDF file extension Date and time the file was produced Format is yyyy mm ddThh mm ss sssZ Day means entire swath in DayNightFlag Day daylight Both means that part of swath lies in darkness Reprocessed means data has been processed before Processed once means this is the first processing of the data Version of algorithm delivered from the SCF Expect that the product will be reprocessed again with an improved algorithm ReprocessingActual LocalVersionID SCF V5 0 0 ReprocessingPlanned reprocessed further update is anticipated This is meaningless information Original pla
26. to the full range of NDSI values 0 0 1 0 Fractional snow is constrained to upper limit of 100 The fractional snow cover map and the snow cover map may be different Fractional snow cover may have greater areal extent because its calculation is not restricted to the same NDSI range as is the snow cover area calculation The fractional snow cover result is screened with the same screens as the snow cover area algorithm Clouds are masked using data from the MODIS Cloud Mask data product MOD35 L2 The MOD35 L2 data is checked to determine if the cloud mask algorithm was applied to a pixel If it was applied then results of the cloud mask algorithm are used If it was not applied then the cloud mask is not used and the 7 of 80 snow algorithm will process for snow assuming that the pixel is unobstructed cloud Only the summary cloud result the unobstructed field of view flag from MODS5 L2 is used to mask clouds in the snow algorithm The day night flag from the MODS35 L2 is also used to mask pixels that lie in night Night is determined where the solar zenith angle is equal to or greater than 85 The snow cover map Snow Cover Reduced Cloud SDS made with selected cloud spectral tests from the cloud mask in Collection 4 is omitted in Collection 5 Though it was possible to reduce cloud obscuration in some situations or reduce cloud commission errors in others those advantages were outweighed by the disadvantage in situations where clouds wh
27. 0 0 SphereCode 1 GridOriginZHDFE UL GROUP Dimension END GROUP Dimension GROUP DataField OBJECT DataField 1 DataFieldName Maximum Snow Extent DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END_OBJECT DataField_1 OBJECT DataField 2 DataFieldName Eight Day Snow Cover DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 2 END GROUP DataField GROUP MergedFields END_GROUP MergedFields END GROUP GRID 1 END GROUPe GridStructure GROUP PointStructure END GROUP PointStructure END Other global attributes in the product are listed in Table 36 Table 36 Other global attributes in MOD10A2 Attribute Name Sample Value Comment HDFEOSVersion HDFEOS V2 9 Version of HDF EOS toolkit used in PGE Number of input days 8 2003 201 2003 202 2003 203 2003 204 2003 205 Days input 2003 206 2003 207 2003 208 Eight day period 2003 201 2003 208 SCF Algorithm Version ld MOD_PR10_AA Internal SCF version of the code modules 56 of 80 Quality Assessment No quality assessment QA data are stored in the product The rationale for QA of the eight day composite product is being discussed Automated QA is not done in the algorithm and the value of passing along the QA data for everyday of input was not a reasonable approach as little was to be gained from that data for the volume that would be used to store it Evaluation and
28. 1 11746 hdf MOD10C1 A2005247 004 2005250144859 hdf MOD10C1 A2005271 004 2005275163947 hdf MOD10C1 A2005272 004 2005276113514 hdf MOD10C1 A2005273 004 2005276193514 hdf EASTBOUNDINGCOORDINATE 180 0 WESTBOUNDINGCOORDINATE 180 0 SOUTHBOUNDINGCOORDINATE 90 0 NORTHBOUNDINGCOORDINATE 90 0 ZONEIDENTIFIER Other Grid System LOCALITYVALUE Global RangeEndingDate 2005 09 30 RangeEndingTime 23 59 59 RangeBeginningDate 2005 09 01 Beginning and ending times for the day Formats are yyyy mm dd RangeBeginningTime 00 00 00 hh mm ss 5 i Version of production generation PGEVersion 5 0 1 executable PGE AssociatedSensorShortName MODIS Sensor name AssociatedPlatformShortName Terra Platform name Instrument and sensor name AssociatedinstrumentShortName MODIS the same Product Specific Attributes PSA QAPERCENTGOODQUALITY 99 Summary quality of data range heck in the algorithm QAPERCENTOTHERQUALITY cyperse cone im he dgarnitnm i Summary percentage of snow SNOWCOVERPERCENT 17 covered land InputPointer Names of MODIS data input files Coverage of entire globe Table 48 Listing of objects in ArchiveMetadata 0 the global attribute in MOD10CM Object Name Typical Value Comment AlgorithmPackageAcceptanceDate 05 2006 AlgorithmPackageMaturityCode Normal Algorithm version information Format is mm AlgorithmPackageName MOD PR10A1 mn Algorit
29. 11v05 005 2005071045059 hdf MOD10A1 A2003206 h1 1v05 005 2005071082446 hdf MOD10A1 A2003207 h11v05 005 2005071122905 hdf MOD10A1 A2003208 h1 1v05 005 2005071161501 hdf MODAPS 2005 03 12T23 26 05 000Z none 40 0 30 0 69 27241 91 37851 IRIX64 mtvs3 6 5 10070055 IP35 5 0 Algorithm version information Format is mm yyyy Descriptive name of the product May be displayed as the product name in the EOS Data Gateway or other dataset search tools Long name of MODIS Names of MODIS input files MODIS Adaptive Processing System Date of processing Format is yyyy mm ddThh mm ss sssZ Archaic and meaningless Eastern western northern and southern most points of the swath Format is decimal degrees Processing done in either UNIX or Linux environment Descriptor file associated with the PGE The StructMetadata 0 global attribute is used by the HDF EOS toolkit to create the mapping relationships between the defined grid and data SDSs Parameters of the projection are stored StructMetadata 0 Listing of StructMetadata 0 for MOD10A2 StructMetadata 0 GROUP SwathStructure END GROUP SwathStructure GROUP GridStructure GROUP GRID_1 GridName MOD Grid Snow 500m XDim 2400 55 of 80 YDim 2400 UpperLeftPointMtrs 7783653 637667 4447802 078667 LowerRightMtrs 667 1703 1 18000 3335851 559000 Projection GCTP_SNSOID ProjParams 6371007 181000 0 0 0 0 0 0 0 0 0 0
30. 122905 hdf MOD10A1 A2003208 h1 1v05 005 2005071 161501 hdf GringPointLongitude 80 765781 91 37851 78 110572 69 036814 These latitude and longitude pairs of points when GringPointLatitude 29 845932 40 0 40 053954 29 891994 connected in a clockwise series form a polygon of the GringPointSequenceNo 1 2 3 4 Names of MODIS data input InputPointer files 53 of 80 swath coverage Always represents the outer ring of coverage ExclusionGRingFlag N RangeEndingDate 2003 07 27 RangeEndingTime 23 59 59 RangeBeginningDate 2003 07 27 Beginning and ending times of the first and last scan line in inninaTi Q0 10 00 the swath Formats are yyyy RangeBeginningTime 00 10 00 im dd hh min ss Version of production Pee version 91 generation executable PGE AssociatedSensorShortName MODIS Sensor name AssociatedPlatformShortName Terra Platform name Instrument and sensor name AssociatedInstrumentShortName MODIS arethe same Product Specific Attributes PSA QAPERCENTGOODQUALITY 100 Summary quality of data range checks done in the QAPERCENTOTHERQUALITY 0 algorithm EM N 11 direction 0 35 VERTICALTILENUMBER LN In latitude direction In latitude direction 0 17 17 i Format is pshhhvvv p projection code TilelD 51011005 S size 1 is full size hhh horizontal tile number vvv vertical tile number CSNOWc
31. 50 00 000000 2 Version of production generation PGEVersion 5 0 4 executable PGE AncillarylnputPointer 991943 005 2000035 1122TA Name of the geolocation file AncillarylnputType Geolocation einer ancilary gata referenced y AssociatedSensorShortName MODIS Sensor name AssociatedPlatformShortName Terra Platform name re th AssociatedlnstrumentShortName MODIS RE SENEO NAmE Are ME Product Specific Attributes PSA QAPERCENTGOODQUALITY 100 Summary quality of data range checks QAPERCENTOTHERQUALIT 0 done in the algorithm Y GRANULENUMBER 239 Unique granule identifier SNOWCOVERPERCENT 03 EUR percentage of snow covered The ArchiveMetadata 0 global attribute contains information relevant to version of the algorithm production environment and geographic location of the data product Contents are described in Table 9 Beginning and ending times of the first and last scan line in the swath Formats are yyyy mm dd hh mm ss ssssss Table 9 Listing of objects in ArchiveMetadata 0 the global attribute in MOD10_L2 Object Name Typical Value Comment AlgorithmPackageAcceptanceDate 05 2006 AlgorithmPackageMaturityCode Normal AlgorithmPackageName MOD_PR10 AlgorithmPackageVersion 5 Algorithm version information Format is mm 19 of 80 Descriptive name of the product May be displayed LongName MODS Terra Snow Cover 5 Min L2 Swath 500m as the product nam
32. AT32 DimList Coarse swath lines 5km Coarse swath pixels 5km END OBJECT 2GeoField 2 END GROUP GeoField GROUP DataField OBJECT DataField 1 DataFieldName Snow Cover DataType DFNT UINT8 21 of 80 DimList Along swath lines 500m Cross swath pixels 500m END OBJECT DataField 1 OBJECT DataField 2 DataFieldName Snow Cover Pixel QA DataType DFNT UINT8 DimList Along swath lines 500m Cross swath pixels 500m END OBJECT DataField 2 OBJECT DataField 3 DataFieldName Fractional_ Snow Cover DataType DFNT UINT8 DimList Along swath lines 500m Cross swath pixels 500m END OBJECT DataField 3 END GROUP DataField GROUP MergedFields END_GROUP MergedFields END GROUP SWATH 1 END GROUP SwathStructure GROUP GridStructure END GROUP GridStructure GROUP PointStructure END GROUP PointStructure END The other global attributes in the product are listed in Table 10 Table 10 Other global attributes in MOD10 L2 Attribute Name Sample Value HDFEOSVersion HDFEOS V2 9 L1BcalibrationQuality marginal L1BmissionPhase EXECUTION L1BnadirPointing Y L1BversionID 2003 07 17 SCF Algorithm Version 919 9 MOD 10 AA Surface Temperature Screen Threshold 283 0 HDFEOS FractionalOffset Along swath lines 500m MOD Swath Snow 0 500000 HDFEOS FractionalOffset Cross swath pixels 500m MOD Swath Snow 0 000000 22 of 80 Comment Version of HDF EOS toolkit used in PGE Quality indicators of MODO2HKM
33. Errors Snow errors from the MOD10A1 inputs are propagated into the eight day product The origin of the errors is snow cloud confusion from the MOD10 L2 product Snow errors of commission are typically manifest as snow in locations and seasons where snow is impossible or very unlikely As the algorithm was designed to map maximum snow cover with no filtering for snow errors the error present is the maximum error in snow extent for the period Errors from every day which probably occur in different locations on different days are mapped which increases the spatial extent of error in the snow map Screening of snow errors is possible in some situations by using the maximum snow cover data and eight day snow cover data together Typically the snow errors associated with cloud shadows and snow cloud confusion occur in different places on different days typically they do not persist in the same location over an eight day period If the assumption that snow errors exist on single days and that snow exists on two or more days is made Single day snow errors may be screened by removing snow that was observed on only a single day in the period A single day occurrence in the eight day snow cover data is indicated when the value is equal to two of a power 0 7 That type of screen may work in the summer but pose problems in transition seasons or winter when single day snow cover may actually exist Other options may be to limit analysis to certain geographi
34. GS GCTP parameters Brief descriptions of the snow data products are given here to give perspective to the sequence Expanded descriptions of the snow products are given in following sections The first product MOD10 L2 has snow cover maps snow extent and fractional snow maps at 500 m spatial resolution for a swath The snow maps are the result of the algorithm identifying snow and other features in the scene Geolocation data latitude and longitude at 5 km resolution are stored in the product The second product MOD10L2G is a multidimensional data product 3 of 80 created by mapping the pixels from the MOD10 L2 granules for a day to the appropriate Earth locations on the sinusoidal map projection thus multiple observations i e pixels covering a geographic location cell in the tile are stacked on one another all snow maps are included Information on the pixels mapped into a cell is stored in pointer and geolocation products associated with the L2G product The third product MOD10A1 is a tile of daily snow cover maps at 500 m spatial resolution The daily observation that is selected from multiple observations in a MOD10L2G cell is selected using a scoring algorithm to select the observation nearest local noon and closest to nadir The fourth product MOD10C1 is a daily global snow cover map in a geographic map projection It is created by assembling MOD10A1 daily tiles and binning the 500 m cell observations to the 0 05 spatial
35. MODIS Snow Products User Guide to Collection 5 George A Riggs Dorothy K Hall Vincent V Salomonson November 2006 Introduction The Snow User Guide to Collection 5 of the MODIS snow products has been infused and expanded with information regarding characteristics and quality of snow products at each level A user should find information on characteristics and quality that affect interpretation and use of the products content this guide includes information and explanations that should enlighten a user s understanding of the products Each product section of the guide has been expanded to include descriptions and explanations of characteristics and quality of the product and the online guide has links or future links to imagery and graphics exemplifying those characteristics The MODIS snow product suite is created as a sequence of products beginning with a swath scene and progressing through spatial and temporal transformations to a monthly global snow product Each snow product in the sequence after the swath product assimilates accuracy and error from the preceding product A user must understand how the accuracy and quality of that daily snow product is affected by the previous level s of input products Distribution statistics from the DAAC reveal that the daily tile snow product is the most frequently distributed of the snow products Review of the literature also shows that the daily and eight day products are the most utili
36. NGULARSIZE CHARACTERISTICBINSIZE GEOANYABNORMAL GEOESTMAXRMSERROR DATACOLUMNS DATAROWS GLOBALGRIDCOLUMNS GLOBALGRIDROWS AlgorithmPackageAcceptanceDate AlgorithmPackageMaturityCode AlgorithmPackageName AlgorithmPackageVersion LongName InstrumentName LocallnputGranulelD Processing Center Typical Value 15 0 463 312716527778 False 50 0 2400 2400 86400 43200 05 2006 Normal MOD PR10A1 p MODIS Terra Snow Cover Daily L3 Global 500m SIN Grid Moderate Resolution Imaging SpectroRadiometer MOD10L2G A2003201 h09v04 005 2006043032816 hdf MODMGGAD A2003201 h09v04 005 2006043030423 hdf MODPTHKM A2003201 h09v04 005 2006043030339 hdf MODO9GHK A2003201 h09v04 005 2006043031 930 hdf MOD120Q1 A2001001 h09v04 004 2004358134052 hdf MODAPS 32 of 80 Comment Estimated maximum error in geolocation of the data in meters Columns in tile Rows in tile Columns across global grid Rows across global grid Algorithm version information Format is mm yyyy Descriptive name of the product May be displayed as the product name in the EOS Data Gateway or other dataset search tools Long name of MODIS Names of MODIS input files MODIS Adaptive Processing System ProcessingDateTime SPSOParameters NorthBoundingCoordinate SouthBoundingCoordinate EastBoundingCoordinate WestBoundingCoordinate 2006 02 12T03 40 28 000Z none 49 9999999955098 39
37. OVERPERGENT HIS percentage of CSNOWcOVERPERGENT bo RE covered land The ArchiveMetadata 0 global attribute contains information relevant to version of the algorithm production environment and geographic location of the data product Contents are described in Table 35 Table 35 Listing of objects ArchiveMetadata 0 the global attribute in MOD10A2 Object Name Typical Value Comment CHARACTERISTICBINANGULARSIZE 15 0 CHARACTERISTICBINSIZE 463 312716527778 DATACOLUMNS 2400 Columns in tile DATAROWS 2400 Rows in tile GLOBALGRIDCOLUMNS 86400 Columns across global grid GLOBALGRIDROWS 43200 Rows across global grid Obieotname eU comcs 54 of 80 AlgorithmPackageAcceptanceDate AlgorithmPackageMaturityCode AlgorithmPackageName AlgorithmPackageVersion LongName InstrumentName LocallnputGranulelD Processing Center ProcessingDateTime SPSOParameters NorthBoundingCoordinate SouthBoundingCoordinate EastBoundingCoordinate WestBoundingCoordinate Processing Environment DESCRRevision 01 2005 Normal MOD_PR10A2 5 MODIS Terra Snow Cover 8 Day L3 Global 500m SIN Grid Moderate Resolution Imaging SpectroRadiometer MOD10A1 A2003201 h11v05 005 2005070055251 hdf MOD10A1 A2003202 h1 1v05 005 2005070125403 hdf MOD10A1 A2003203 h11v05 005 2005070195037 hdf MOD10A1 A2003204 h1 1v05 005 2005071010128 hdf MOD10A1 A2003205 h
38. S at satellite reflectance image from swath of MODO2HKM for 3 January 2003 A Snow cover appears as yellow in this display of bands 1 4 and 6 Snow cover map of the swath B and the snow cover map in sinusoidal projection C Night No Decision 75 of 80 Figure 2 MODIS snow cover from swath of 10 L2 for 3 January 2003 1745 GMT in Fractional snow cover map in B and fractional snow map in sinusoidal projection in C FRACTIONAL SNOW Hox 1 20 B 21 40 1 50 51 60 Wl ct 70 71 80 E 81 90 191 100 c ovo mask nicht water lt Nater Lake Ice Cloud Ocean Night No Decision 76 of 80 Figure 3 MODIS snow cover map and corresponding snow cover pixel map B from swath of MOD10 L2 for 3 January 2003 1745 GMT No Decision OBSCURED B 1 5 mo Il o 70 89 190 100 nicut water wissinc 77 80 SNOW QA Key GOOD QUALITY OTHER QUALITY CLOUD MASK NIGHT WATER vissinc dl NO DATA Figure 5 Daily global snow map A and cloud obscured map B from MOD10C1 A2000063 3 March 2000 in geographic projection Figure 7 Eight day global snow map and cloud obscured map B from MOD10C2 A2000057 26 Feb 3 March 2000 in geographic projection CLOUD OBSCURED m m mo BH o 3j 70 89 90 100 n
39. UP DataField OBJECT DataField 1 DataFieldName Snow Cover Daily Tile DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE 33 of 80 DeflateLevel 9 END OBJECT DataField 1 OBJECT DataField 2 DataFieldName Snow Spatial QA DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 2 OBJECT DataField 3 DataFieldName Snow Albedo Daily Tile DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 3 OBJECT DataField 4 DataFieldName Fractional_ Snow Cover DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 4 END GROUP DataField GROUP MergedFields END_GROUP MergedFields END_GROUP GRID_1 END GROUP GridStructure GROUP PointStructure END GROUP PointStructure END The other global attributes in the product are listed in Table 20 Table 20 Other global attributes in MOD10A1 Attribute Name HDFEOSVersion L2GAutomaticQualityFlag Sample Value Comment Version of HDF EOS toolkit MORBOS vee used in PGE Passed L2G Quality indicators L2GAutomaticQualityFlagExplanation Output file is created and good Method of calculating pixel L2GCoverageCalculationMethod volume coverage a grid cell Number of swaths covering L2GNumberOfOverlapGranules 4 some part of the tile L2GFirstLayerSelectionCriteria order of inpu
40. algorithm affecting a relatively small percentage of the data Analysis of erroneous snow the mapping of features not snow as snow has revealed causes for erroneous snow Causes corrections and solutions to alleviate erroneous snow mapping are presented in the following subsections Warm Bright Surface Features In the first processing of MODIS data it was discovered that some surface features e g salt pans or sandy beaches were being mapped as snow because they had reflectance characteristics similar to snow specifically the NDSI value of those features was similar to snow Mismatch of the land water mask used in processing to the geolocated MODIS data was and still is a problem The majority of that erroneous snow occurred in climatically warm regions of the world where snow was not likely to occur in any season The solution to this type of snow error was to apply a thermal screen to remove the error The surface temperature algorithm was taken from the sea ice algorithm integrated into the snow algorithm and used as a screen to prevent very warm snow like pixels from being mapped Any pixel identified as snow but that has an estimated temperature 2283 is changed to land This screening is a rough estimate of surface temperature as the surface temperature is calculated as though the pixel is snow covered sea ice That temperature screening was successful at greatly reducing the occurrence of erroneous snow in warm regions of the world an
41. ality of data display During the summer 37 of 80 season some coastal regions mainly the Antarctic Peninsula may be snow free for a brief period of time Study of such areas should use the MOD10 L2 or MOD10A1 products A mask of where occurrence of snow is extremely unlikely e g the Amazon the Sahara Great Sandy Desert is applied at the end of the algorithm to eliminate erroneous snow occurrence Source of erroneous snow in those regions is the MOD10_L2 product where erroneous snow detection occurs and is carried forward through the processing levels to the CMG At the CMG level the use of this extremely unlikely snow mask eliminates erroneous snow from the masked regions but will allow it in regions where snow may be a rare event Scientific Data Sets Day_CMG_Snow_Cover The percentage of snow covered land observed in the CMG cell is given in the Day_CMG_Snow_Cover SDS Fig 7a Snow cover percentage is the fraction of snow covered land observed based on the entire amount of land mapped in the CMG grid cell No attempt was made to interpret snow cover possibly obscured by cloud Percentage of snow is reported in the range of 0 100 Table 23 Local attributes for Day Snow Cover Attribute name long name units format coordsys valid range FillValue Mask value Night value Cell resolution Water mask la nd threshold 76 Definition Long Name of the SDS SI units of the data if any How the
42. ame MOD CMG Snow 5km XDim 7200 YDim 3600 UpperLeftPointMtrs 180000000 000000 90000000 000000 LowerRightMtrs 180000000 000000 90000000 000000 46 of 80 Projection GCTP_GEO GridOriginZHDFE GD UL GROUP Dimension END GROUP Dimension GROUP DataField OBJECT DataField 1 DataFieldName Day CMG Snow Cover DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 1 OBJECT DataField 2 DataFieldName Day CMG Confidence Index DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 2 OBJECT DataField 3 DataFieldName Day Cloud Obscured DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 3 OBJECT DataField 4 DataFieldName Snow Spatial QA DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 4 END GROUP DataField GROUP MergedFields END_GROUP MergedFields END GROUP GRID 1 END GROUP GridStructure GROUP PointStructure END GROUP PointStructure END The other global attributes in the product are listed in Table 29 47 of 80 Table 29 Other global attributes MOD10C1 Attribute Name Sample Value Comment HDFEOSVersion HDFEOS_V2 9 Version of HDF_EOS toolkit used in PGE MOD10A2 Snow cover over eight days is mapped as maximum snow extent Fig 8 and as a chronology of snow observa
43. an y the data valid range Max and min values within a 0 254 selected data range 10 of 80 Data used to fill gaps in the _ FillValue wath 255 0 100 fractional snow 200 missing data 201 no decision 211 night 225 land 237 inland water 239 250 cloud 254 detector saturated 255 fill Key to meaning of data in the Key SDS Nadir_data_res Nominal pixel resolution at nadir 500 m olution HDF predefined attribute names Latitude and Longitude Coarse resolution 5 km latitude and longitude data for geolocating the snow data are located in the Latitude and Longitude SDSs The latitude and longitude data correspond to a center pixel of a 5 km by 5 km block of pixels in the snow SDSs The mapping relationship of geolocation data to the snow data is specified in the global attribute StructMetadata 0 Mapping relationship was created by the HDF EOS SDPTK toolkit during production Geolocation data is mapped to the snow data with an offset 5 and increment 10 The first element 1 1 in the geolocation SDSs corresponds to element 5 5 in Snow_Cover SDS the algorithm then increments by 10 in the cross track or along track direction to map geolocation data to the Snow_Cover SDS elements Local attributes are listed in Table 5 and Table 6 Table 5 Local attributes with Latitude SDS Attribute name Definition Value Coarse 5 km long_name Long Name of the SDS resolulion latitude units SI units of
44. ble observation of all the swath level observations mapped into a grid cell for the day using the scoring algorithm Fractional snow is reported in the 0 100 range including inland water bodies Pixels that are not snow are labeled as water cloud or other condition A color coded image of a fractional snow map is shown in Figure 6b HDF predefined and custom local attributes are stored The HDF predefined attributes may be used by some software packages The custom local attributes are specific to the data in the SDS Local attributes are listed in Table 15 25 of 80 Value Snow cover extent by best observation of the day none I3 cartesian 0 254 255 O missing data 1 no decision 11 night 25 no snow 37 lake 39 ocean 50 cloud 100 lake ice 200 snow 254 detector saturated 255 fill Table 15 Local attributes with Fractional Snow Cover SDS Attribute name long name units format coordsys valid range FillValue Key Definition Long Name of the SDS SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Key to meaning of data in the SDS HDF predefined attribute names Snow Albedo Daily Tile The snow albedo algorithm result is stored as a map of the snow albebo for the tile The snow albedo map correspond
45. bration Support Team MCST web page and in supporting documentation If missing data is encountered those pixels are identified as missing data in MOD10 L2 If unusable data is encountered then a no decision result is written 6 of 80 for those pixels Usable L1B calibrated radiance data is converted to at satellite reflectance for use in the snow algorithm Snow covered area is determined through the use of two groups of grouped criteria tests for snow reflectance characteristics in the visible and near infrared regions and screening of snow decisions Global criteria for snow is a normalized snow difference index NDSI 4 band 6 band 4 band 6 greater than 0 4 and near infrared reflectance band 2 greater than 0 11 and band 4 reflectance greater than 0 10 If a pixel passes that group of criteria tests it is identified as snow The minimum reflectance tests screen low reflectance surfaces e g water that may have a high NDSI value from being erroneously detected as snow To enable detection of snow in dense vegetation a criteria test using NDSI and the normalized difference vegetation index NDVI of band 2 band 1 band 2 band 1 is applied to pixels that have an NDSI value in the range of 0 1 to 0 4 In this criteria test a pixel with NDSI and NDVI values in a defined polygon of a scatter plot of the two indices and that has near infrared reflectance in band 2 greater than 0 11 and band 1 reflectance greater than 0 1 i
46. c regions of interest in a tile which may allow better logic for screening snow errors or to find persistent snow cover during the period Reduction of snow errors will occur as a result of reducing the snow errors in the MOD10 L2 product MOD10C2 The eight day climate modeling grid snow cover data product is generated by merging all the MOD10A2 products tiles for an eight day period Table 22 and binning that 500 m data to 0 05 or about 5 6 km resolution to create a global CMG map of maximum snow extent Fig 9 Snow extent cloud cover confidence index and quality assessment data are included in the product Algorithm Description The MOD10A2 500 m resolution data are mapped into the corresponding cell of the CMG Approximately 120 observations go into each CMG cell Input values are binned into categories of snow cloud night etc The percentages of 57 of 80 snow percentage of cloud QA confidence index are computed based on the binning results for each cell of the CMG and written into the appropriate SDSs The basis for the percentage calculations is the amount of land in that cell determined from the base land extent map A land base map used in binning the MOD10A2 data was created from the University of Maryland 1 km global land cover mask http glcf umiacs umd edu data landcover index shtml The base land extent map indicates the amount of land in a CMG cell and is used to determine if the cell is proces
47. cription of the mapping relationships is not given here Use of HDF EOS toolkit other EOSDIS supplied toolkits DAAC tools or other software packages may be used to geolocate the data or to transform it to other projections and or data file formats Listing of objects in the global attribute StructMetadata 0 in MOD10 L2 StructMetadata 0 GROUP SwathStructure GROUP SWATH_1 SwathName MOD Swath Snow GROUP Dimension OBJECT Dimension 1 DimensionName Along swath lines 500m Size 4060 END OBJECT Dimension 1 OBJECT Dimension 2 DimensionName Cross swath pixels 500m 20 of 80 2708 END OBJECT Dimension 2 OBJECT 2Dimension DimensionName Coarse swath lines 5km 406 END OBJECT Dimension 3 OBJECT Dimension 4 DimensionName Coarse swath pixels 5km 271 END OBJECT Dimension 4 END GROUP Dimension GROUP DimensionMap OBJECT DimensionMap 1 GeoDimension Coarse swath pixels 5km DataDimension Cross swath pixels 500m Offset 5 Increment 10 END OBJECT DimensionMap 1 OBJECT DimensionMap 2 GeoDimension Coarse swath lines 5km DataDimension Along swath lines 500m Offset 5 Increment 10 END OBJECT DimensionMap 2 END GROUP DimensionMap GROUP IndexDimensionMap END_GROUP IndexDimensionMap GROUP GeoField OBJECT GeoField_1 GeoFieldName Latitude DataType DFNT_FLOAT32 DimList Coarse_swath_lines_5km Coarse_swath_pixels_5km END OBJECT GeoField 1 OBJECT GeoField 2 GeoFieldName Longitude DataType DFNT FLO
48. d along warm coastal regions especially those with wide sandy beaches Coastline Differences The land water mask the MODIS geolocation product MODO3 is used to control processing path in the snow algorithm In Collection 5 the land water mask stored in the geolocation product was developed by the MODIS science team at Boston University That land water mask contains many improvements over the previously used MODIS land water mask Accuracy of coastlines and location of water bodies is improved Yet misalignment of coastlines with the geolocated MODIS swath data still exists That misalignment causes erroneous snow mapping to occur along coastlines in several regions around the world This problem is readily apparent in the Canadian Arctic Islands in the summer when the islands may have a snowy coastline During the summer the Canadian Arctic islands appear to have snowy coastlines in places obviously in error During the transition seasons no snow error is apparent because snow 14 of 80 cover is expected those seasons During boreal winter darkness no error is seen The snow error appears seasonal but is year round because of the land water mask to image misalignment of coastlines Snow error on coastlines in warm regions is usually removed by the temperature screen but not always The mixed signal of ocean beach and coastline misalignment remains as a minor problem Inland Water Bodies as Snow or Lake Ice In the new BU land wa
49. data Version of the L1B processing algorithm Internal SCF version of the code modules Temperature K setting for this screen Offset for better geolocation of data Offset for better geolocation of data MOD10 L2G Snow Product L2G product is the result of mapping all the MOD10 L2 swaths acquired during a day to grid cells of the Sinusoidal map projection The Earth is divided into an array of 36 x 18 longitude by latitude tiles about 10 x10 in size in the Sinusoidal projection The MODL2G algorithm was created as a generic gridding algorithm for many of the MODIS data products in the land discipline group and was customized to each MODIS data product as necessary See Wolfe et al 1999 for a description of the gridding technique and product contents The L2G algorithm maps pixels from the MOD10 L2 SDSs into cells of the projection No calculations or analysis of snow is done at L2G The MOD10 L2G and other L2G products are necessary intermediate products used as input to the daily snow cover algorithm generating the MOD10A1 product The MOD10_L2G is not archived at the DAAC thus is not available for order through ECS The other L2G products are archived at a DAAC and can be ordered MOD10A1 The daily snow product is a tile of data gridded in the sinusoidal projection Tiles are approximately 1200 x 1200 km 10 x10 in area Snow data arrays produced by selecting the most favorable observation pixel
50. data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans For seasonal darkness Nominal grid cell resolution Decision point to process a cell as land or water 38 of 80 Value Daily snow extent global at 5km none I3 latitude longitude 0 100 255 254 111 0 05 deg 12 00000 Antarctica sno w note Key Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names Day CMG Cloud _Obscured The percentage cloud obscuration for a cell is given in the Day CMG Cloud Obscured SDS Fig 7b The percentage of cloud is the count of cloud observations for the day based on the total number of land cells in the grid cell That is the same basis as used to calculate the percentage of snow A cell may range from clear 0 cloud to completely cloud obscured 100 cloud Antarctica deliberately mapped as snow 0 100 of snow in cell 107 lake ice 111 night 250 cloud obscured water 253 data not mapped 254 water mask 255 fill Table 24 Local attributes for Day CMG Cloud Obscured Attribute name long name units format coordsys valid range FillValue Mask value Definition Long Name of the SDS SI units of the data if any How the data should be viewed Fortran f
51. e data should be viewed IMS Fortran format notation s Coordinate system to use for coordsys ihadafa latitude longitude valid ande Max and min values within a 0 1 fang selected data range _FillValue Data used to fill gaps in the 255 swath Cell resolution Nominal grid cell resolution 0 05 deg Antarctica Antarctica sno Antarctica masked as perennial deliberately w note Snow cover mapped as snow O other quality 1 good quality Key Key to meaning of data in the 252 Antarctica SDS mask 254 water mask 255 fill HDF predefined attribute names Snow Map Accuracy and Errors Analysis of the quality of the MOD10CM has been limited to visual and qualitative comparative analysis of the monthly fractional snow maps Prior to Collection 5 processing the MOD10CM generated in Collection 4 processing was available only by request from the Few if any reports regarding analysis or evaluation of the MOD10CM appear in the literature to the present Overall the MOD10CM appears to be a reasonable estimate of the mean monthly fractional snow cover when compared to other sources of global or regional snow maps Validation status is Stage 1 but may change as evaluation and validation analysis is done on the product Global Attributes There are five global attributes in the MOD10CM product three are ECS defined CoreMetadata 0 ArchiveMetadata 0 and StructMetadata 0 and the others are product defined These global attributes s
52. e for the purpose of aiding a user in understanding and interpreting the data product The snow algorithm is described in detail in the Algorithm Theoretical Basis Document ATBD Analysis for snow in a MODIS swath is done on pixels of land or of inland water that have nominal L1B radiance data are in daylight and the cloud mask is applied A snow decision is also screened for temperature and difference of a band ratio to reduce the occurrence of erroneous snow in some situations Data inputs to the snow algorithm are listed in Table 2 Land and inland waters are masked with the 1 km resolution land water mask contained in the MODIS geolocation product MODO3 In Collection 5 the land water mask made by the Boston University BU team based on EOS data is used During Collection 4 the BU land water mask replaced the EOS land water mask that had been used More information is given on the land water mask in QA sections below The 1 km data of the land water mask is applied to the four corresponding 500 m pixels in the snow algorithm Ocean waters are not analyzed for snow Inland waters lakes and rivers are analyzed for snow covered ice conditions The MODIS L1B is screened for missing data and for unusable data Unusable data results from the processing at L1B when the sensor radiance data fails to meet acceptable criteria MODIS data may be unusable for several reasons Specifics of L1B processing and criteria can be found at the MODIS Cali
53. e in the EOS Data Gateway or other dataset search tools InstrumentName Moderate Resolution Imaging SpectroRadiometer Long name of MODIS MODO2HKM A2003198 1945 005 2006036052819 hdf MOD021KM A2003198 1945 005 2006036052819 hdf Names of MODIS input MOD35 L2 A2003198 1945 005 20060360701 11 hdf files MOD03 A2003198 1945 005 20060351 12242 hdf LocallnputGranulelD MODIS Adaptive Processing Center MODAPS Processing System Date of processing Format ProcessingDateTime 2006 02 05T15 01 35 000Z is yyyy mm ddThh mm ss sssZ SPSOParameters none Archaic and meaningless EastBoundingCoordinate 58 9066026791 133 Eastern western northern WestBoundingCoordinate 176 825688181697 and southern most points of the swath Format is NorthBoundingCoordinate 86 7594955695887 decimal degrees SouthBoundingCoordinate 61 6178586242137 Linux minion5009 2 6 8 1 24mdksmp 1 SMP Thu Jan Processing Environment 13 23 11 43 MST 2005 i686 Intel R Xeon TM CPU 2 40GHz unknown GNU Linux Processing done in either UNIX or Linux environment Descriptor file associated DESCRRevision 5 0 with the PGE The StructMetadata 0 global attribute is created by the HDF EOS toolkit to specify the mapping relationships between the geolocation data and the snow cover data SDSs referred to as data fields in the structural metadata Mapping relationships are unique in HDF EOS and are stored in the product using HDF structures Des
54. ere not mapped as clouds and thus as land by the snow algorithm though it was actually snow covered land beneath the clouds It is possible to make selective use of the cloud mask spectral tests and other data for snow mapping however refinement of that approach was not pursued for Collection 5 Table 2 MODIS data product inputs to the MODIS snow algorithm ESDT Long Name Data Used Reflectance for MODIS bands MODIS Level 1B Calibrated and 1 0 645 um NUDO EM Geolocated Radiances 2 0 865 um 4 0 555 um 6 1 640 um MODIS Level 1B Calibrated and 31 11 28 Geolocated Radiances 32 12 27 um Land Water Mask Solar Zenith Angles MODO03 MODIS Geolocation Sensor Zenith Angles Latitude Longitude Cloud Mask Flag Unobstructed Field of View Flag Day Night Flag MODS5 L2 MODIS Cloud Mask Scientific Data Sets Snow Cover Results of the snow cover mapping algorithm are stored as coded integers in the Snow Cover SDS The snow cover algorithm identifies pixels as snow snow covered water body typically lakes or rivers land water cloud or other condition A color coded image of a snow map is shown in Figure 1b a winter image of the northern US plains and south central Canada alongside a false 8 of 80 color reflectance image of the swath Fig 1a Images in Fig 1 b are un projected Fig 1c is the snow map in sinusoidal projection HDF predefined and custom local attributes are stored The HDF predefined attribu
55. erve different purposes such as search and order of products mapping and product version tracking and evaluating a product The ECS defined attributes are written as very long character strings in parameter value language PVL format Descriptions of the global attributes are given in the following tables CoreMetadata 0 and ArchiveMetadata 0 are global attributes in which information compiled about the product during product generation is archived 70 of 80 StructMetadata 0 contains information about the swath or grid mapping relevant to the product A user wanting detailed explanations of the global attributes and related information should query the EOSDIS related web sites Table 47 Listing of objects in the global attribute CoreMetadata 0 in MOD10CM Object Name LocalGranulelD ProductionDateTime DayNightFlag ReprocessingActual LocalVersionID ReprocessingPlanned ScienceQualityFlag AutomaticQualityFlagExplanation AutomaticQualityFlag ScienceQualityFlagExplanation QAPercentMissingData QAPercentCloudCover ParameterName VersionID ShortName Sample Value MOD10CM A2005244 005 2005283201645 hdf 2005 10 10T20 16 45 000Z Day reprocessed SCF V5 0 0 further update is anticipated Not investigated No automatic quality assessment done in the PGE Passed See http landweb nascom nasa gov cgi bin QA WWW qaFlagPage cgi sat terra the product Science Quality status
56. extent map The maximum snow cover extent map is generally reasonable if limited to 8096 or greater snow percentage levels and occurrence of persistent cloud is accounted for Snow errors of commission probably dominate the lower e g less than 2096 snow cover level in many situations Global Attributes There are four global attributes in the MOD10C2 product three are ECS defined CoreMetadata 0 ArchiveMetadata 0 and StructMetadata 0 and the others are product defined These global attributes serve different purposes such as search and order of products mapping and product version tracking and evaluating a product The ECS defined attributes are written as very long character strings in parameter value language PVL format Descriptions of the global attributes are given in the following tables CoreMetadata 0 and ArchiveMetadata 0 are global attributes in which information compiled about the product during product generation is archived StructMetadata 0 contains information about the grid mapping relevant to the product A user wanting detailed explanations of the global attributes and related information should query the EOSDIS related web sites Table 41 Listing of objects in the global attribute CoreMetadata 0 in MOD10C2 Object Name Sample Value Comment Filename of product Format is EDST Ayyyyddd vvv yyyydddhhmmss hdf Ayyyyddd acquisition date vvv collection version yyyydddhhmmss date and time of production
57. f a lot of cloud cover and that snow percentage may not be a good estimate because of the cloud cover obscuring all or parts of a cell A simplified example will be used to demonstrate the calculations for percent snow percent cloud and confidence index A 5 km 0 05 CMG grid cell has 50 500m observations distributed as follows snow observations 20 snow free land observations 15 cloud obscured observations 10 other but not water observations 5 The percent snow is computed as Snow 100 Number of snow observations number of cloudless land and other land observations Snow 100 20 20 15 10 5 Snow 40 The percent cloud is computed as Cloud 100 Number of cloud observations number of cloudless land and other land observations Cloud 100 10 20 15 10 5 36 of 80 Cloud 20 The confidence index Cl is computed as Cl 100 Number of clear land observations number of cloudless land and other land observations Cl 100 20 15 20 15 10 5 CI 70 A number of possible snow cloud and land combinations and the CI calculated for them are listed in Table 22 The highest is always associated with clear view conditions at any percentage of snow cover When clouds completely obscure the surface the is 0 because the surface is not seen In situations where there are only snow and cloud observations in a cell the Cl will be the same as the percent snow
58. fc nasa gov MODIS Land Discipline http modis land gsfc nasa gov Cloud Mask MOD35 http cimss ssec wisc edu modis1 pdf CMUSERSGUIDE PDF 8 MODIS Characterization Support Team http www mcst ssai biz mcstweb 9 MODIS Atmosphere Discipline http modis atmos gsfc nasa gov 10 MODAPS Services http modaps nascom nasa gov services O C1 HDF EOS Information and Tools 11 EOSDIS http spsosun gsfc nasa gov ESDIShome html 12 HDF http Awww hdfgroup org 13 HDF EOS http ndfeos gsfc nasa gov Note Samples of HDF EOS files can be obtained from this site 14 ECS Data Handling System http edhs1 gsfc nasa gov 15 MODIS Data Support http daac gsfc nasa gov MODIS software shtml other 16 HEG Tool HDF EOS to GIS format conversion tool http eosweb larc nasa gov PRODOCS misr tools geotiff tool html Earth Science 17 GSFC Earth Sciences Portal http earthsciencesportal qsfc nasa qov 80 of 80
59. generates the snow and cloud cover maps based the total number of observations of a class e g snow cloud snow free land etc and total number of land observations mapped into a cell of the CMG Observations from all the input cells of the MOD10A1 corresponding to a CMG cell approximately 3600 per CMG cell at the equator are put in observation bins Calculated snow maps are stored as SDSs in the MOD10C1 product The objective of the algorithm and resulting product is to provide the user an estimate of snow cover extent that was observed in a CMG cell along with an estimate of how much of the land surface was obscured by clouds and an index that estimates the confidence in the estimates Table 21 MODIS data product inputs to the MOD10C1 snow algorithm ESDT Long Name Data Used MODIS Terra Snow Cover Daily L3 MOD10A1 500m SIN Grid Snow cover The binning algorithm places the different classes of observations e g snow lake cloud etc into bins for each class A land bin is used in MOD10C1 algorithm to sum all observations made of land e g snow snow free land cloud 35 of 80 over land etc That sum of land counts is the basis for expressing the percentage of snow cloud and the confidence index for each CMG cell A CMG specific land base mask was made for use with the binning algorithm The 0 05 land mask was derived from the University of Maryland 1km global land cover data set http glcf umiacs umd edu data landco
60. granule ordered The post production QA metadata may or may not be present depending on whether or not the data granule has been investigated The xml file should be examined to determine if 5 of 80 postproduction QA has been applied to the granule The Quality Assessment sections of this guide provide information on postproduction QA The data products were generated in the ECS science data production system using the HDF EOS Version 5 2 9 Science Data Processing SDP Toolkit HDF and the C programming language Various software packages commercial and public domain are capable of accessing the HDF EOS files MOD10 L2 The swath product is generated using the MODIS calibrated radiance data products MODO2HKM and MODO21 KM the geolocation product MODO3 and the cloud mask product MOD35 L2 as inputs The MODIS snow cover algorithm output product MOD10 L2 contains two SDS of snow cover a quality assessment QA SDS latitude and longitude SDSs local attributes and global attributes The snow cover algorithm identifies snow covered land snow covered ice on inland water and computes fractional snow cover There are approximately 288 swaths of Terra orbits acquired in daylight so there are approximately 288 MOD10 L2 snow products per day An example of the MOD10 L2 product snow cover map is exhibited in Figure 1a c in both un projected and projected formats Algorithm Description A sketch of the snow algorithm is given her
61. hdf HDF file extension LocalGranulelD MOD10C2 A2003201 005 2005072123100 hdf 194 Date and time the file was ProductionDateTime 2005 03 13T12 31 00 000Z produced Format is yyyy mm ddThh mm ss sssZ Day means entire swath in daylight Both means that part of swath lies in darkness DayNightFlag Both ReprocessingActual reprocessed 63 of 80 Reprocessed means data has been processed before Processed once means this is the first processing of the data LocalVersionID SCF V5 0 0 ReprocessingPlanned Version of algorithm delivered from the SCF Expect that the product will be reprocessed again with an improved algorithm further update is anticipated This is meaningless information Original plan was for this metadata to be set updated by investigator after evaluation validation however that plan was dropped and this metadata is not set updated See ScienceQualityFlagExplanati on for current information ScienceQualityFlag Not investigated No automated QA checks made during execution of the algorithm AutomaticQualityFlagExplanation No automatic quality assessment done in the PGE Default setting because no automated QA checks are done AutomaticQualityFlag Passed See http landweb nascom nasa gov cgi bin QA WWW qaFlagPage cgi sat terra the product Science Quality status Amount of data missing f
62. he MOD10 L2 product MOD10CM Monthly global snow extent data product has been added to the sequence of MODIS snow products for both Terra and Aqua General The bit encoded spatial quality assessment data has been replaced with an integer spatial quality assessment data value A local attribute named Key has been included with all SDSs This is the key to meaning of data values in the data array 2 of 80 naming convention for the SDS was implemented so there is greater naming consistency through the data products Some SDS names are different in Collection 5 New in Collection 5 is the use of HDF internal compression in the level 3 and higher products to reduce the volume of the data files in the archive and the amount of network resources required to transport the data files The internal compression should be invisible to users and software packages that can read the HDF HDF EOS format For the advanced user the internal compression does create Vgroup and Vdata within the product The level 2 swath products are compressed using the NCSA HDF hrepack command line compression tool instead of internal compression coding which may or may not be invisible depending on software used to access the data products It may be necessary to uncompress the data using hrepack See http hdf ncsa uiuc edu tools hrepack hrepack html for information and usage Sequence of Snow Products Snow data products are produced as a series of seven produc
63. hiving user services geolocation and analysis of data The ECS global attributes are written in parameter value language PVL and are stored as a character string Metadata and values are stored as objects within the PVL string Products may also contain product specific attributes PSAs defined by the product developers as part of the ECS CoreMetadata 0 attribute Geolocation and gridding relationships between HDF EOS point swath and grid structures and the data are contained in the ECS global attribute StructuralMetadata 0 Other information about mapping algorithm version processing and structure may be stored in the ArchiveMetadata 0 also in PVL or as separate global attributes Other information about the product may be stored in global attributes separate from the ECS global attributes Stored with each SDS is a local attribute that is a key to the data values in the SDS There may also be other local attributes with information about the data Detailed descriptions of the SDSs are given for each snow product in following sections A separate file containing metadata will accompany data products ordered from a DAAC That metadata file will have an xml extension and is written in Extendable Markup Language The xml file contains some of the same metadata as in the product file but also has other information regarding archiving and user support services as well as some post production quality assessment QA information relevant to the
64. hmPackageVersion 5 MODIS Terra Snow Cover Monthly L3 Global 0 05Deg CMG Descriptive name of the product May be displayed as the product name in the EOS Data Gateway or other dataset search tools Moderate Resolution Imaging InstrumentName SpectroRadiometer Long name of MODIS PLATFORMSHORTNAME Terra LongName 72 of 80 GLOBALGRIDCOLUMNS 7200 GLOBALGRIDROWS 3600 Processing Center MODAPS MODIS Adaptive Processing System si is Date of processing Format is yyyy mm ProcessingDateTime 2005 10 10T16 16 33 000000Z daThhanmae sss SPSOParameters none Archaic and meaningless DESCRRevision 5 0 Descriptor file associated with the PGE IRIX64 mtvs1 6 5 10070055 IP35 Processing done in either UNIX or Linux Processing Environment environment DESCRRevision 5 0 Descriptor file associated with the PGE The StructMetadata 0 global attribute is created by the HDF EOS toolkit to specify the mapping relationships between the map projection data and the snow cover data SDSs Mapping relationships are unique in HDF EOS and are stored in the product using HDF structures Description of the mapping relationships is not given here Use of HDF EOS toolkit other EOSDIS supplied toolkits DAAC tools or other software packages may be used to map the data or to transform it to other projections Map projection parameters are from the GCTP Listing of the global attribute StructMetadata 0 in MOD10CM St
65. icht water 1 120 sow we 78 of 80 Figure 8 Monthly global average snow cover map for March 2006 i Bow 3ow oE WE 60 E 90 120 E 150 References Hall D K and G A Riggs 2006 submitted Assessment of errors in the MODIS suite of snow cover products Hydrological Processes Klein and Stroeve J 2002 Development and validation of a snow albedo algorithm for the MODIS instrument Annals of Glaciology vol 34 pp 45 52 Salomonson V V and Appel 2004 Estimating the fractional snow covering using the normalized difference snow index Remote Sensing of Environment 89 3 351 360 Salomonson and Appel 2006 Tekeli A E Sensoy A Sorman A Aky rek Z and Sorman 2006 Accuracy assessment of MODIS daily snow albedo retrievals with in situ measurements in Karasu basin Turkey Hydrol Process 20 705 721 Wolfe R E D P Roy E Vermote 1999 MODIS land data storage gridding and compositing methodology level 2 grid IEEE TGARS July 1999 36 4 pp1324 1338 http modis snow ice gsfc nasa gov atbd html 79 of 80 Related Web Sites EOS 1 Terra Website http terra nasa gov Aqua Website http aqua nasa gov 2 ECS http ecsinfo gsfc nasa gov 3 National Snow and Ice Data Center http nsidc org MODIS 4 MODIS Snow Ice Global Mapping Project http modis snow ice gsfc nasa gov MODIS Project http modis gs
66. idence in the extent of snow Cloud obstruction reduces the confidence index Table 38 Local attributes for Eight Day Confidence Index Attribute name Definition Value Confidence index long name Long Name of the SDS for the eight day 59 of 80 units format coordsys valid range FillValue Mask value Water mask la nd threshold 76 Antarctica confi dence index n ote Key SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans Decision point to process a cell as land or water Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names Eight Day CMG Cloud Obscured The cloud obscured data indicates how much of the land surface in the cell was persistently obscured during the eight day period snow map none I3 latitude longitude 0 100 255 254 12 00000 Antarctica deliberately mapped as snow Confidence index set to 100 0 100 confidence index value 107 lake ice 111 night 250 cloud obscured water 253 data not mapped 254 water mask 255 fill Table 39 Local attributes for Eight Day Cloud _Obscured Attribute name long name Definition Long Name of the SDS 60 of 80 Value Dloud obscuration units
67. ithm did not use solar zenith as a factor in scoring the observations A new scoring algorithm that included solar zenith as a factor in scoring the observations was implemented in V004 on 13 September 2004 That new algorithm effectively chose observations from near local solar noon thus eliminating the problem of erroneous snow caused by low illumination from the MOD10A1 thus increasing its quality and in turn quality of the MOD10C1 snow map However the erroneous snow problem remained in the MOD10 L2 product Snow and Cloud Confusion Snow and cloud discrimination problems persist in the algorithm that result in typically very small amounts of erroneous snow mapped in some cloud situations This error is associated with parts of ice clouds which appear yellow in a MODIS band 1 4 6 color display The error occurs on parts of the clouds that lie in the shadow of other parts of the cloud or on parts that have a middling amount of reflectance This problem is associated with these types of clouds and can occur in any season in about any location Analysis has been focused on North America The problem exhibits greatest impact on quality in summer when these cloud types situations are more frequent and result in erroneous snow mapping The amount of snow error attributable to these snow cloud situations is usually very small in terms of pixel counts in the 0 001 to 0 196 range but may range up to about 396 depending on extent type and pattern of cloud
68. l snow data The single QA SDS applies to both the snow cover area and fractional snow cover SDSs Snow_Cover_Pixel_QA The quality assessment data provides an indication of the quality of the input data for the snow and fractional snow algorithms Data for a pixel are determined to be of good quality other quality or may be set to a thematic value for certain conditions Unless the input data is unusable or missing the data 12 of 80 quality will usually be set to good example of the snow cover pixel QA is shown in Figure 3 Local attributes are listed in Table 7 Table 7 Local attributes with Snow Cover Pixel QA SDS Attribute name Definition Value Snow cover per long name Long Name of the SDS pixel thematic QA units SI units of the data if any none How the data should be viewed TORIA Fortran format notation p x Coordinate system to use for i coordsys ihe data Cartesian x Max and min values within a valid range selected data range grana FillValue Data used to fill gaps in the 255 swath 0 good quality 1 quality 252 Antarctica Key Key to meaning of data in the mask 253 land SDS mask 254 mask saturated 255 fill HDF predefined attribute names Indicators of quality are also given in metadata objects in the CoreMetadata 0 global attribute generated during production or in post product scientific and quality checks of the data product Of the few quality metadata object
69. n the global attribute CoreMetadata 0 in MOD10 L2 Object Name Sample Value Comment LocalGranulelD MOD10 L2 A2003198 1945 005 200603615004 Filename of product Format is 17 of 80 3 hdf EDST Ayyyyddd hhmm vvv yyyydddhh mmss hdf Ayyyyddd hhmm acquisition date and time in UTC Vmeormesesunr ae and time the file was produced ProductionbateTime 0 Umeeormesesz 00000 07 16 05 52 31 0002 Format is yyyy mm ddThh mm ss sssZ Day means entire swath in daylight Both means that part of swath lies in darkness DayNightFlag Day Reprocessed means data has been processed before Processed once means this is the first processing of the data Version of algorithm delivered from the SCF Expect that the product will be reprocessed again with an improved algorithm ReprocessingActual LocalVersionID ReprocessingPlanned reprocessed SCF V5 0 4 further update is anticipated This is meaningless information Original plan was for this metadata to be set updated by investigator after evaluation validation however that plan was dropped and this metadata is not set updated See ScienceQualityFlagExplanation for current information i No automatic quality assessment done in the No automated QA checks made during AutomaticQualityFlagExplanation PGE execution of the algorithm R Default setting because no automated AutomaticQualityFlag
70. n was for this metadata to be set updated by investigator after evaluation validation however that plan was dropped and this metadata is not set updated See ScienceQualityFlagExplanatio n for current information ScienceQualityFlag Not investigated No automated QA checks made during execution of the algorithm AutomaticQualityFlagExplanatio n No automatic quality assessment done in the PGE Default setting because no automated QA checks are done AutomaticQualityFlag Passed See http landweb nascom nasa gov cgi bin QA_WWW qgaFlagPage cgi sat terra the product Science Quality status 0 Amount of data missing from the input file Amount of land in the swath QAPercentCloudCover 0 obscured by clouds Mavi QA parameters given apply to ParameterName Maximum Snow Extent the snow cover data URL where updated information on science QA should be posted ScienceQualityFlagExplanation QAPercentMissingData Indicates the EOSDIS VersionID 5 Collection ShortName MOD10A2 ESDT name of product MOD10A1 A2003201 h11v05 005 2005070055251 hdf iil MOD104A1 A2003202 h11v05 005 2005070125403 hdf MOD104A1 A2003203 h11v05 005 20050701 95037 hdf MOD10A1 A2003204 h11v05 005 2005071010128 hdf MOD10A1 A2003205 h11v05 005 2005071045059 hdf MOD10A1 A2003206 h1 1v05 005 2005071082446 hdf MOD10A1 A2003207 h11v05 005 200507 1
71. nd evaluating a product The ECS defined attributes are written as very long character strings in parameter value language PVL format Descriptions of the global attributes are given in the following tables 43 of 80 CoreMetadata 0 ArchiveMetadata O are global attributes in which information compiled about the product during product generation is archived StructMetadata 0 contains information about the swath or grid mapping relevant to the product A user wanting detailed explanations of the global attributes and related information should query the EOSDIS related web sites Table 27 Listing of objects in the global attribute CoreMetadata 0 in MOD10C1 Object Name LocalGranulelD ProductionDateTime DayNightFlag ReprocessingActual LocalVersionID ReprocessingPlanned ScienceQualityFlag AutomaticQualityFlagExplanation AutomaticQualityFlag ScienceQualityFlagExplanation Sample Value MOD10C1 A2003200 005 2006053045454 hdf 2006 02 22T04 54 54 000Z Both reprocessed SCF V5 0 0 further update is anticipated Not investigated No automatic quality assessment done in the PGE Passed See http landweb nascom nasa gov cgi bin QA_WWW qaFlagPage cgi sat terra the product Science Quality status 44 of 80 Comment Filename of product Format is EDST Ayyyyddd vvv yyyydddhhmmss hdf Ayyyyddd hhmm acquisition date and time in UTC hnnvnn horizontal and vertical
72. ns from Level 1B 0 0 100 0 Band 1 in the swath 0 0 100 0 Saturated EV Obs Band 1 96 The percentage of Saturated observations from Level 1B 0 0 100 0 Band 2 in the swath 0 0 100 0 Saturated EV Obs Band 2 96 The percentage of saturated observations from Level 1B 0 0 100 0 Band 4 in the swath 0 0 100 0 Saturated EV Obs Band 4 96 The percentage of saturated observations from Level 1B 0 0 100 0 Band 6 in the swath 0 0 100 0 HDF predefined attribute names Saturated EV Obs Band 6 96 Fractional Snow Cover Results of the fractional snow cover algorithm are stored as coded integers in the Fractional Snow Cover SDS The fractional snow algorithm calculates fractional snow in the 0 100 range including inland water bodies Pixels that are not identified as snow are labeled as water cloud or other condition A fractional snow map is shown in Figure 2 HDF predefined and custom local attributes are stored The HDF predefined attributes may be used by some software packages The custom local attributes are specific to the data in the SDS Local attributes are listed in Table 4 Table 4 Local attributes with Fractional Snow Cover SDS Attribute name Definition Value Fractional snow long name Long Name of the SDS cover 500m units SI units of the data if any none How the data should be viewed format l Fortran format notation m Coordinate system to use for cartesi
73. oblem appears to be related to a cloud spectral visible reflectance test in the cloud mask algorithm that gives a fairly confident result of cloud so the pixel is mapped as cloud Investigation of the problem has been sporadic as it is a low priority compare to other snow problems and a possible solution to the problem has not been formulated though investigation done suggests that individual cloud spectral test s and processing path flags may need to be read to better understand and possibly solve the problem specific to snow mapping Global Attributes There are 11 global attributes in the MOD10 L2 product three are ECS defined CoreMetadata 0 ArchiveMetadata 0 and StructMetadata O and the others are specific to the product These global attributes serve different purposes such as search and order of products mapping product version tracking and evaluating a product The ECS defined attributes are written as very long character strings in parameter value language PVL format Descriptions of the global attributes are given in the following tables CoreMetadata 0 and ArchiveMetadata 0 are global attributes in which information compiled about the product during product generation is archived StructMetadata 0 contains information about the swath or grid mapping relevant to the product A user wanting detailed explanations of the global attributes and related information should query the EOSDIS related web sites Table 8 Listing of objects i
74. ociated with the PGE The StructMetadata 0 global attribute is created by the HDF EOS toolkit to specify the mapping relationships between the geolocation data and the snow LongName 65 of 80 cover data 5055 Mapping relationships are unique HDF EOS and stored in the product using HDF structures Description of the mapping relationships is not given here Use of HDF EOS toolkit other EOSDIS supplied toolkits DAAC tools or other software packages may be used to geolocate the data or to transform it to other projections Map projection parameters are from the GCTP Listing of the global attribute StructMetadata 0 in MOD10C2 StructMetadata O StructMetadata O GROUP SwathStructure END _GROUP SwathStructure GROUP GridStructure GROUP GRID_1 GridName MOD_CMG_Snow_5km XDim 7200 YDim 3600 UpperLeftPointMtrs 180000000 000000 90000000 000000 LowerRightMtrs 180000000 000000 90000000 000000 Projection GCTP_GEO GridOriginZHDFE GD UL GROUP Dimension END GROUP Dimension GROUP DataField OBJECT DataField 1 DataFieldName Eight Day Snow Cover DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 1 OBJECT DataField 2 DataFieldName Eight Day Confidence Index DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 2 OBJECT DataField 3 DataFieldName Eight Day Cloud Ob
75. ormat notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans 39 of 80 Value Daily cloud obscuration percentage none I3 latitude longitude 0 100 255 254 Not processed value Night value Cell resolution Water mask la nd threshold 76 Antarctica clou d note Key For seasonal darkness Nominal grid cell resolution Decision point to process a cell as land or water Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names Day_CMG_Confidence_Index An index of the confidence in the snow observation being a good or poor estimate of snow cover in a cell is stored in this SDS The Cl ranges from 0 100 252 111 0 05 deg 12 00000 Antarctica deliberately mapped as snow Cloud value set to 252 0 100 percent of cloud in cell 107 lake ice 1112night 250 cloud obscured water 253 data not mapped 254 water mask 255 fill Table 25 Local attributes for Day CMG Confidence _ Index Attribute name long name units format Definition Long Name of the SDS SI units of the data if any How the data should be viewed Fortran format notation 40 of 80 Value Confidence index for the daily snow map none coordsys valid range FillValue Mask value Cell resolution
76. over A daily cell must have a Confidence Index of 7096 to be included in the average That filter is applied so that only the clearest of the daily observations are included in the average See the MOD10C1 section for description of the A daily observation contributes to the monthly average for a cell as follows Daily contribution to monthly mean 100 snow Cl For daily observations that are cloud free the snow contribution to the mean is the observed snow fraction For daily observations of mixed snow and cloud fractions with a high Cl it is assumed that there is some fraction of snow cover obscured by cloud In that case the daily snow observation is increased in 67 of 80 that equation so that the contribution to the monthly mean will be greater than the daily snow observation For example cell has 25 snow cover and the Cl 75 then the cell is determined to have 2596 75 100 33 fractional snow cover Daily observations with a Cl lt 70 are assigned either as 100 cloudy night missing or no decision There must be at least one day in the month for each cell with the Cl gt 70 in order for the mean snow cover to be computed for that cell of the monthly CMG If that restriction is not met then the cell is reported as no decision A second filter is applied to the calculated mean fractional snow cover of each cell to filter out those cells in which the magnitude of snow cover is less than 1096 Cells failing the
77. pretation relevant to their application Because of the poor quality of the snow products over Antarctica the continent is masked as perennial snow cover in the daily snow CMG product That poor quality originates in the MOD10 L2 algorithm and is caused by the great difficulty in discriminating between clouds and snow over Antarctica Masking was done to increase visual quality of the image but excludes Antarctica from scientific analysis To reduce erroneous snow mapping in regions of the world that climatologically should never have snow a snow not possible mask was created and applied in the algorithm The effect has been to eliminate erroneous snow in many parts of the world Those erroneous snow errors were caused by either deeply shadowed surfaces or snow cloud confusion errors on some types of clouds The mask is spatial all seasonal climatology so snow would not be possible in these areas during any season Along some coasts some snow may appear as the snow impossible map and the product map are not perfectly aligned Those errors originate with land water mask misalignments from MOD10 L2 and passed forward to this level Global Attributes There are 11 global attributes in the MOD10A1 product three are ECS defined CoreMetadata 0 ArchiveMetadata 0 and StructMetadata 0 and the others are specific to the product These global attributes serve different purposes such as search and order of products mapping and product version tracking a
78. rcomputing Applications NCSA is the standard archive format for EOS Data Information System EOSDIS products The snow product files contain global attributes metadata and scientific data sets SDSs i e data arrays with local attributes Unique in HDF EOS data files is the use of HDF features to create point swath and grid structures to support geolocation of data The geolocation information and relationships between data in a SDS and geographic coordinates latitude and longitude or map projections to support mapping the data supporting mapping stored as Vgroup and Vdata in the file The SDSs are attached as data fields to the HDF EOS swath or grid structure The geolocation data can only be accessed from the StructMetadata 0 attribute In order to geolocate the data the StructMetadata 0 must be accessed to get geographic information and the data fields i e SDSs attached to it for mapping It is possible to access the SDSs without having to access the StructMetadata O but the geolocation information will not be attached to the SDS Users unfamiliar with HDF and HDF EOS formats may wish to consult web sites listed in the Related Web Sites section for more information Snow data product files contain three EOS Data Information System EOSDIS Core System ECS global attributes also referred to as metadata by ECS These ECS global attributes CoreMetadata 0 ArchiveMetadata 0 and StructMetadata 0 contain information relevant to production arc
79. re beyond the scope of this user guide but are discussed in the MODIS snow ATBD modis snow ice gsfc nasa gov Despite the different band usage the snow map algorithms are very similar and the quality of snow mapping is very similar though subtle differences exist between the products The higher level Level 3 product algorithms are the same for Terra and Aqua Similarities and differences between Terra and Aqua are presented in the appropriate product section The guide is organized into overview sections and data product sections Overview sections cover commonalities in the data products or describe external sources of information relevant to the products Data product sections are composed of a succinct algorithm description data content description and explanations of error and characteristics that should enlighten a user s understanding of each snow product New in Collection 5 Collection 5 reprocessing began in September 2006 starting the first day of MODIS science data acquisition 24 February 2000 Collection 4 data will be available for at least six months after the date that data was reprocessed for Collection 5 MOD10 L2 Fractional snow cover area has been added as a data array in the swath product for both Terra and Aqua The snow cover map with reduced cloud approach has been deleted from the data product MOD10A1 A fractional snow cover data array has been added to the product Fractional snow cover data is input from t
80. rom QAPercentMissingData 0 the swath Amount of land in the swath QAPercentCloudCover 4 obscured by clouds E QA parameters given apply ParameterName Eight Day Global Snow Cover to the snow cover data URL where updated information on science QA should be posted ScienceQualityFlagExplanation Indicates the EOSDIS VersionID 5 Collection ShortName MOD10C2 MODAPSops3 PGE AM1 M coeff PGE67 MOD_PR10C2 cmgTL5km global anc hdf MOD10A2 A2003201 h16v00 005 2005072085912 hdf MOD10A2 A2003201 h17v00 005 2005072085941 hdf MOD10A2 A2003201 h18v00 005 200507209001 7 hdf MOD10A2 A2003201 h22v15 005 2005072092707 hdf MOD10A2 A2003201 h23v15 005 2005072092707 hdf MOD10A2 A2003201 h24v15 005 2005072092707 hdf EASTBOUNDINGCOORDINATE 180 0 WESTBOUNDINGCOORDINATE 180 0 SOUTHBOUNDINGCOORDINATE 90 0 NORTHBOUNDINGCOORDINATE 90 0 ZONEIDENTIFIER Other Grid System LOCALITYVALUE Global RangeEndingDate eve tie RangeEndingTime 23 59 59 ESDT name of product Lil Names of MODIS data input InputPointer files Coverage of entire globe 64 of 80 Beginning and ending times Do II H the day Formats are 222 2003 07 20 RangeBeginningTime 0 0000007 00005 5955 0502520 00 00 yyyy mm dd hh mm ss Version of production PGEVersion 5 0 2 generation executable PGE AssociatedSensorShortName MODI 0505 5 55052525 5
81. roneous snow but leave actual snow e g snow covered mountains 42 of 80 unaffected or minimally so In transition seasons and winter erroneous snow is likely to be more difficult to screen because snow is expected in those seasons however there is indication errors like this occur less during the winter season Analysis into possible seasonality affected occurrence of erroneous snow has not been undertaken Data from the snow cover and cloud obscured SDSs and CI could be used together to better understand the reported fractional snow observation For example if a completely snow covered region was viewed and no clouds obstructed the view on that day then percentage of snow cover would be 100 If that snow covered region was viewed but there was 30 cloud obscuration that day then percentage of snow cover would be 7096 A user could use the cloud obscured data for the cell to determine that there was 3096 cloud obscuration for that day and could use the Cl to make an interpretation that only clouds and snow were observed in the cell From that information it would be possible to make an interpretation if desired about snow cover existing or not under the cloud cover In situations of partially snow covered and snow free land with partial cloud cover the snow cloud and Cl could be used to make an interpretation of snow cover on the ground despite the partial cloud cover A user is encouraged to make best use of combinations of the data for inter
82. ructMetadata 0 StructMetadata O GROUP SwathStructure END GROUP SwathStructure GROUP GridStructure GROUP GRID 1 GridName MOD CMG Snow 5km XDim 7200 YDim 3600 UpperLeftPointMtrs 180000000 000000 90000000 000000 LowerRightMtrs 180000000 000000 90000000 000000 Projection GCTP_GEO GridOriginZHDFE UL GROUP Dimension END GROUP Dimension GROUP DataField OBJECT DataField 1 DataFieldName Snow Cover Monthly DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END_OBJECT DataField_1 OBJECT DataField 2 DataFieldName Snow Spatial QA 73 of 80 DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 2 END GROUP DataField GROUP MergedFields END_GROUP MergedFields END GROUP GRID 1 END GROUP GridStructure GROUP PointStructure END GROUP PointStructure END The other global attributes in the product are listed in Table 49 Table 49 Other global attributes in MOD10CM Attribute Name Sample Value HDFEOSVersion HDFEOS_V2 9 MOD10C1 A2005244 004 2005247012647 hdf MOD10C1 A2005246 004 2005249111746 hdf MOD10C1 A2005247 004 2005250144859 hdf InputFileNames MOD10C1 A2005273 004 200527619351 4 hdf 74 of 80 MOD10C1 A2005271 004 2005275163947 hdf MOD10C1 A2005272 004 2005276113514 hdf Comment Version of HDF EOS toolkit used in PGE Listing of the MOD10C1 input files Figures Figure 1 MODI
83. rvations MODES Daily L2G Global 500m SIN Grid coverage observation swath and location MODIS Terra Surface Reflectance Surface reflectance MODOSGHK L2G Global 500m SIN Grid bands 1 5 and 7 MOD12Q1 MODIS Terra Land Cover Type Yearly Land cover type L3 Global 1km SIN Grid Scientific Data Sets Snow_Cover_Day_Tile The snow cover map is the result of selecting the most favorable observation of all the swath level observations mapped into a grid cell for the day Mapped is snow snow covered water bodies typically lakes or rivers land water cloud or other condition A color coded image of a snow map is shown in Figure 6a HDF predefined and custom local attributes are stored The HDF predefined attributes may be used by some software packages The custom 24 of 80 local attributes are specific to the data in the SDS Local attributes listed in Table 14 Table 14 Local attributes for Snow Cover Day Tile Attribute name long name units format coordsys valid range FillValue Key Definition Long Name of the SDS SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Key to meaning of data in the SDS HDF predefined attribute names Fractional Snow Cover The fractional snow cover map is the result of selecting the most favora
84. s The source of error lies with those clouds not being mapped as certain cloud by the cloud mask because the clouds do not dominate the reflectance of 16 of 80 the 1 km resolution cloud mask When those missed clouds are processed in the snow algorithm they appear to have spectral features particularly the NDSI that are more like snow than a not snow feature The snow algorithm processes those pixels as not cloud and the NDSI signal being similar to snow causes the pixels to be identified as snow Snow and cloud confusion errors of this type have been noticeably reduced in Collection 5 due to improvement of the cloud mask algorithm which currently detects these types of clouds more often thus classifying them correctly as cloud and preventing them from being analyzed erroneously as snow However great the improvement there still remains albeit a very small amount of shaded yellow cloud that is not identified as cloud and is then mapped as snow in the snow algorithm Snow as Cloud At the edges of snow cover in the mountains or on plains the edge of the snow is frequently identified as cloud by the cloud mask algorithm This problem is sometimes very obvious extending over several kilometers of sparse or thin snow at edge of a snow cover Sometimes the problem is not so obvious occurring as only a pixel or two width in the mountains lf there is a sharp boundary between deep snow and snow free land the problem may not occur This pr
85. s determined to be snow This latter criteria test is applied without regard to the ecosystem Snow covered ice on inland water is determined by applying the global criteria for snow detection to pixels mapped as inland water by the land water mask Another screen is applied to the snow decision of all the above criteria tests to reduce erroneous snow detections A surface temperature screen of 283 K is applied to prevent bright warm surfaces from being erroneously detected as snow The screen functions to reduce the occurrence of erroneous snow detection in some situations and is described in subsections of the Quality Assessment section Intermediate checks for theoretical bounding of reflectance data and the NDSI ratio are made in the algorithm In theory reflectance values should lie within the 0 100 range and the NDSI ratio should lie within the 1 0 to 1 0 range Summary statistics are kept within the algorithm for pixels that exceed these theoretical limits however the test for snow is done regardless of violations of these limits These violations suggest that error or other anomalies may have crept into the input data and indicate that further investigation may be warranted to uncover the causes Fractional snow cover is computed for all land and inland water body pixels in a swath Fractional snow cover is calculated using the regression equation of Salomonson and Appel 2004 and in press The fractional snow cover calculation is applied
86. s in the CoreMetadata 0 global attribute only the ScienceQualityFlagExplanation is relevant as a pointer to website for science quality status No automatic quality assessment is done in the algorithm production nor is science quality checked during production Snow Accuracy and Errors Under ideal conditions of illumination clear skies and several centimeters of snow a smooth surface the snow algorithm is about 93 100 accurate at mapping snow Hall and Riggs submitted Ideal conditions are usually not the norm so the snow algorithm was designed to identify snow globally in nearly any situation The NDSI has proved to be a robust indicator of snow around the globe The NDSI is a reliable indicator of snow when snow is present Patchy snow or thin snow cover on vegetated surfaces may be missed by the NDSI 13 of 80 Experience and analysis of MODIS snow products over three collections of data have revealed strengths and weaknesses in the snow mapping technique Originally the snow algorithm was designed to map snow globally and was unrestricted in global application Robustness of the snow mapping algorithm is exhibited in the relatively rare errors of missing snow when snow is present That approach maximized ability to detect snow and had the consequence of also increasing errors of commission identifying non snow features as snow in the snow cover algorithm Mapping features as snow erroneous snow is a persistent problem with the snow
87. s to snow mapped in the snow cover map in Snow Cover Day Tile SDS Snow albedo is reported in the 0 100 range and non snow features are also mapped using different data values A color coded image of a snow albedo map is shown in Figure 6c HDF predefined and custom local attributes are stored The HDF predefined attributes may be used by some software packages The custom local attributes are specific to the data in the SDS Local attributes are listed in Table 16 Value Fractional snow covered land for the tile none I3 cartesian 0 254 255 0 100 fractional Snow 200 missing data 201 decision 211 night 225 land 237 inland water 239 ocean 250 cloud 254 detector saturated 255 fill Table 16 Local attributes with Snow Albedo Daily Tile SDS Attribute name long name Definition Long Name of the SDS 26 of 80 Value Snow albedo of the corresponding Snow cover observation units SI units of the data if any none tama How the data should be viewed Fortran format notation Coordinate system to use for y the data Max min values within 1 valle tange selected data range 97109 2 Data used to fill gaps in the FillValue 255 missing value Value for missing data 250 0 1002snow albedo 101 decision 111 night 125 land 137 inland water 139 Key to meaning of data in the 150 cloud SDS iar 250 missing 251 self
88. s will exist In warm regions or warm seasons in temperate regions of the world the coastal snow errors that might be caused by land water mask misalignment are usually corrected by the thermal screen in MOD10 L2 thus do not appear or may have a seasonal appearance depending on the region Though the MOD10A1 product is generated for Antarctica it is considered of very poor quality on the continent because of the great difficulty in identifying cloud cover and discriminating between cloud and snow there A very obvious problem occurs when cloud is present but not identified as cloud by the cloud mask algorithm In that situation the snow algorithm assumes a cloud free view and either identifies the surface as not snow covered or identifies the cloud as snow In either case the result is wrong Such confusion occurs fairly frequently especially in coastal regions and is exhibited as patches of snow free Antarctica surface In MOD10A1 algorithm no action is taken to resolve the problem thus the problem is available for investigation In the higher level snow products e g MOD10C1 Antarctica is masked as 100 snow cover to eliminate the snow errors and generate a good visual product there but one that is not useful for scientific study Validation and evaluation of the snow albedo data is ongoing Snow albedo is estimated to be within 1096 of surface measured snow albedo based on studies in the literature Klein and Stroeve 2002 Tekeli et al 2006 and
89. scured DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END OBJECT DataField 3 OBJECT DataField 4 66 of 80 DataFieldName Snow Spatial QA DataType DFNT UINT8 DimList YDim XDim CompressionType HDFE COMP DEFLATE DeflateLevel 9 END_OBJECT DataField_4 END GROUP DataField GROUP MergedFields END_GROUP MergedFields END GROUP GRID 1 END GROUP GridStructure GROUP PointStructure END GROUP PointStructure END The other global attributes in the product are listed in Table 43 Table 43 Other global attributes in MOD10C2 Attribute Name Sample Value Comment HDFEOSVersion HDFEOS_ V2 9 Version of HDF_EOS toolkit used in PGE MOD10CM This product is a global 0 05 resolution monthly mean fractional snow cover extent derived from MODIS daily snow cover extent CMG MOD10C1 products for a month Fig 10 The monthly mean fractional snow cover is generated using all the days of a month Algorithm Description The algorithm computes the average fractional snow cover for each cell in the CMG using the 28 31 days of MOD10C1 for the month Data is filtered so that the most relevant days of snow cover are used to calculate the average and to filter out data that is of low magnitude i e low occurrence of snow during the month The later filter works to remove some occurrences of erroneous snow from the monthly snow average The daily snow data is used to compute the monthly average snow c
90. sed for snow and is used in the calculation of the confidence index A land percentage of 1296 in a CMG cell is used as the threshold to determine that a cell is considered as land Antarctica is arbitrarily mapped as perennial snow cover because Antarctica is 99 or greater snow covered During the summer up to 1 may be snow free mostly on the Antarctic Peninsula Mapping Antarctica as always snow covered was done to eliminate the errors of snow detection or snow cloud discrimination that occur in the MOD10 L2 algorithm from being passed into the product A night condition polar darkness is handled by determining the latitude of the CMG cell nearest the equator that is full of night observations All CMG cells poleward from that latitude are mapped as night Night was handled that way so that a neat demarcation of night and day is shown in the CMG A mask of where occurrence of snow is extremely unlikely e g the Amazon the Sahara Great Sandy Desert is applied at the end of the algorithm to eliminate erroneous snow occurrences Source of erroneous snow is the MOD10 L2 product where false snow detection occurs and is carried forward through the processing levels At the CMG level the use of this extremely unlikely snow mask eliminates erroneous snow from selected regions but will allow for snow detection in regions where snow may be a rare event There are four SDSs with local attributes and four global attributes written in the CMG prod
91. t pointer MOW Ihe inputs were Staged in time order from 00 00 to 34 of 80 23 59 MOD10_L2 A2003201 1710 005 2006036191945 hdf MOD10_L2 A2003201 1845 005 2006036194834 hdf List of MOD10_L2 swaths MOD10_L2 A2003201 2020 005 2006036192728 hdf mapped into the tile MOD10_L2 A2003201 2025 005 2006036192626 hdf MOD10InputGranuleNames Internal SCF version of the SCF Algorithm Version ld MOD PR10 code modules MOD10C1 The daily global climate modeling grid CMG a geographic projection snow product gives a global view of snow cover at 0 05 resolution Fig 7 Snow cover extent is mapped by processing the MOD10A1 products approximately 320 tiles of land data for a day into the CMG Snow cover extent is expressed as a percentage of snow observed in a grid cell of the CMG at 0 05 resolution based on the MOD10A1 cells at 500 m mapped into a grid cell A corresponding map of cloud cover percentage is also generated and stored The snow and cloud percentage arrays can be used together to get a comprehensive view of snow and cloud extents for a day Since the cells of the CMG may contain mixed features an expression of confidence in the extent of snow is determined and stored along with other QA data Algorithm Description A binning algorithm is used to calculate snow cover cloud cover confidence index and quality assessment in a 0 05 CMG cell based on the 500 m MOD10A1 input data Table 21 The binning algorithm
92. ter mask some but not all inland water bodies rivers lakes etc are included in the land water mask The BU group provides these insights to the land water mask A great amount of interpretation was involved in the mapping of these water bodies Though it would seem a distinctly easy task to make a distinction between water and land it was very difficult Difficulties were encountered gauging the size of water bodies due to turbidity conditions amount of vegetation in water bodies and in boreal regions confusion between snow and ice and lack of frequent clear sky views for mapping The end result is that water bodies exceeding 1 km in dimension are included Water bodies less than 1 km in size were not included in the mask Also water bodies of isolated single pixels in extent were excluded As a result many inland rivers are discontinuous or absent Features on the land water mask toward the polar regions may be distorted and coastlines may display shearing due to the way the land water mask was generated and projected Missing water bodies are likely to have an effect on reflectance in the pixel s in which they occur The snow algorithm uses the land water mask to direct the processing path to land or inland water body For small water bodies differences between the land water mask and what is imaged can lead to errors in the snow map in both classes of snow and ice covered lakes Errors at shores of larger water bodies may also occur as a result
93. tes may be used by some software packages The custom local attributes are specific to the data in the SDS Local attributes are listed in Table 3 Table 3 Local attributes with Snow Cover SDS Attribute name long name units format coordsys valid range FillValue Key Nadir data res olution Valid EV Obs Band 1 96 Valid EV Obs Band 2 96 Valid EV Obs Band 4 96 Definition Long Name of the SDS SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Key to meaning of data in the SDS Nominal pixel resolution at nadir The percentage of valid observations from Level 1B in Band 1 in the swath 0 0 100 0 The percentage of valid observations from Level 1B in Band 2 in the swath 0 0 100 0 The percentage of valid observations from Level 1B in 9 of 80 Value Snow covered land none I3 cartesian 0 254 255 O missing data 1 no decision 11 night 25 no snow 37 lake 39 ocean 50 cloud 100 lake ice 200 snow 254 detector saturated 255 fill 500 m 0 0 100 0 0 0 100 0 0 0 100 0 Band 4 in the swath 0 0 100 0 The percentage of valid observations from Level 1B 0 0 100 0 Band 6 in the swath 0 0 100 0 Valid EV Obs Band 6 The percentage of saturated observatio
94. the data if any degrees Max and min values within a valid range selected data range 90 00 90 00 FillValue Data used to fill gaps in the 999 000 swath MOD03 source Source of data geolocation 11 of 80 product data read from center pixel in 5 km box HDF predefined attribute names Table 6 Local attributes with Longitude SDS Attribute name Definition Value Coarse 5 km long name Long Name of the SDS resolution longitude units SI units of the data if any degrees valid range Max and min values within a 180 00 180 00 selected data range FillValue Data used to fill gaps in the 999 000 swath MOD03 geolocation source Source of data product data read from center pixel in 5 km box HDF predefined attribute names Quality Assessment A revised approach to quality assessment QA was used in Collection 5 Instead of the spatial QA data being bit encoded flags as was done in Collection 4 and prior collections integer numbers are coded to convey the QA information The QA data should be easier to use and gives a general indicator of good or other quality for the data Data quality is determined by making the same checks as in Collection 4 but the result is an integer value stored in the QA SDS The purpose of the spatial QA is to provide information each pixel that can be viewed in the same spatial context as the snow maps The QA data may be used to help determine the usefulness of the snow cover and fractiona
95. thus low values are indicative of extensive cloud cover and high values are indicative of low cloud cover In situations where there is a mix of snow cloud and land the CI is indicative of level of confidence that the reported snow percentage estimates the snow in the cell despite the cloud cover In those situations Cl has higher values with low cloud amounts at any snow amount but the decreases as cloud cover increased indicating decreased confidence in the estimated snow percentage Table 22 Example of how relates to percent snow cover in CMG cell In this example there are a total of 50 input observations cells to the CMG cell All observations are binned as snow snow free land or cloud Snow count Cloud Land count 96 snow 96 cloud CI count 0 0 50 0 0 100 25 0 25 50 0 100 50 0 0 100 0 100 0 25 25 0 50 50 0 50 0 0 100 0 25 25 0 50 50 50 10 40 0 20 80 20 40 10 0 80 20 80 25 10 15 25 10 80 10 25 15 20 50 50 40 5 5 80 5 90 5 5 40 5 5 90 5 35 10 5 70 30 Polar darkness a night condition is handled by determining the latitude of the CMG cell nearest the equator that is full of night observations All CMG cells poleward from that latitude are filled as night Polar darkness is handled this way so that a neat demarcation of night and day is shown in the CMG Antarctica has been masked as perennially snow covered The masking was done to improve the visual qu
96. tions in the two SDS in this product Eight day periods Table 30 begin on the first day of the year and extend into the next year An eight day compositing period was chosen because that is the ground track repeat period of the Terra platform The last eight day period of a year extends into first few days of the next year The product can be produced with two to eight days of input There may not always be eight days of input because of various reasons so the user should check the attributes to determine what days observations were obtained or were missing in a period Table 30 Eight Day Periods Period No Year Days 1 1 8 2 9 16 3 17 24 4 25 32 5 33 40 6 41 48 7 49 56 8 57 64 9 65 72 10 73 80 11 81 88 12 89 96 13 97 104 14 105 112 15 113 120 16 121 128 17 129 136 18 137 144 19 145 152 48 of 80 20 153 160 21 161 168 22 169 176 23 177 184 24 185 192 25 193 200 26 201 208 27 209 216 28 217 224 29 225 232 30 233 240 31 241 248 32 249 256 33 257 264 34 265 272 35 273 280 36 281 288 37 289 296 38 297 304 39 305 312 40 313 320 41 321 328 42 329 336 43 337 344 44 345 352 45 353 360 46 361 368 Includes 2 or 3 days from next year depending on leap year Algorithm Description The algorithm composites eight days of input MOD10A1 to generate maximum snow extent for the period and tracks the chronology of snow observations The multiple days of observations for a cell are examined If snow
97. to data acquisition or production problems The algorithm was designed to run will with fewer than eight days so that the data acquired could be processed even if one to six days of data is unavailable Days used as input are identified in the global attributes Scientific Data Sets Maximum_Snow_ Extent The maximum snow extent for the period depicts where snow was observed on one or more days in the period Fig 8 HDF predefined and custom local attributes are stored The HDF predefined attributes may be used by some software packages The custom local attributes are specific to the data inthe SDS Local attributes are listed in Table 32 Table 32 Local Attributes for the Maximum_Snow_Extent SDS Attribute name Definition Value Maximum snow long name Long Name of the SDS extent over the 8 day period units SI units of the data if any none toma How the data should be viewed Fortran format notation 50 of 80 coordsys valid range FillValue Cell area 2 snow 2 Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Nominal area of cell Estimated area of all cells mapped as snow Key to meaning of data in the SDS HDF predefined attribute names Eight Day Snow Cover Input files are ordered chronologically in the algorithm and for days on which snow was observed a bit in the byte
98. ts The sequence begins as a swath scene at a nominal pixel spatial resolution of 500 m with nominal swath coverage of 2330 km across track by 2030 km along track five minutes of MODIS scans A summarized listing of the sequence of products is given in Table 1 Products in EOSDIS are labeled as Earth Science Data Type ESDT the ESDT label ShortName is used to identify the snow data products The EOSDIS ShortName also indicates what spatial and temporal processing has been applied to the data product Data product levels briefly described Level 1B L1B is a swath scene of MODIS data geolocated to latitude and longitude centers of 1 km resolution pixels A level 2 L2 product is a geophysical product that remains in latitude and longitude orientation of L1B A level 2 gridded L2G product is in a gridded format of a map projection At L2G the data products are referred to as tiles each tile being a piece e g 10 x 10 area of a map projection L2 data products are gridded into L2G tiles by mapping the L2 pixels into cells of a tile in the map projection grid The L2G algorithm creates a gridded product necessary for the level 3 products A level 3 L3 product is a geophysical product that has been temporally and or spatially manipulated and is in a gridded map projection format and comes as a tile of the global grid The MODIS L3 snow products are in the sinusoidal projection or geographic projection Projections are defined using the US
99. uct Scientific Data Sets Eight Day CMG Snow Cover This SDS is the global map of maximum snow cover extent for the eight day period Extent of snow cover observed expressed as percentage of land in the CMG cell is given The valid range of snow cover extent is 0 100956 Table 37 Local attributes for Eight Day CMG Snow Cover Attribute name Definition Value amp Eight day snow long name Long Name of the SDS extent 5km units SI units of the data if any none How the data should be viewed Fortran format notation I3 format 58 of 80 Coordinate system to use for coordsys ihe data latitude longitude valid rande Max and min values within a 0 100 fang selected data range Data used to fill gaps in the _ FillValue Surat 255 Mask value Used for oceans 254 Night value For seasonal darkness 111 Water mask la v nd threshold Decision point to process a cell 12 00000 96 as land or water Antarctica Antarctica sno Antarctica masked as perennial deliberately w note Snow cover mapped as snow 0 100 of snow in cell 107 lake 1112night 237 inland water Ke Key to meaning of data in the 250 cloud y SDS obscured water 253 data not mapped 254 water mask 255 fill HDF predefined attribute names Eight Day CMG Confidence Index The CI indicates how much of the land surface was observed not obscured by clouds The greater the percentage of land observed the higher the conf
100. ue Water mask la nd threshold 76 Antarctica QA note Key Definition Long Name of the SDS SI units of the data if any How the data should be viewed Fortran format notation Coordinate system to use for the data Max and min values within a selected data range Data used to fill gaps in the swath Used for oceans Decision point to process a cell as land or water Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names 62 of 80 Value Snow cover per cell QA none I3 latitude longitude 0 100 255 254 12 00000 Antarctica deliberately mapped as snow QA value set to 1 0 good quality 1 quality 252 Antarctica mask 253 data not mapped 254 mask 255 fill Snow Accuracy and Errors Snow errors from the MOD10A2 inputs are propagated into the eight day CMG product Origin of the errors is the MOD10O L2 product and they have been maximized in extent in the MOD10A2 product An unintended consequence of mapping maximum snow extent was to also maximize the extent of snow errors Since no screens for snow errors were placed in the algorithm the errors propagate between product levels At the eight day CMG level the errors pose a difficulty to using the entire range of snow percentage in all situations However a user may apply screens or filters to reduce the extent of snow errors in the snow cover
101. ver index shtml If a CMG cell contains 1296 or greater land then it is considered land and analyzed if less than 1296 it is considered ocean That threshold was selected as a balance that minimized snow errors along coasts yet was sensitive to mapping snow along coasts The percentage of snow given in cells of the Day CMG Snow Cover SDS is calculated using the 500m data totals of the number of snow observations and count of other land observations in that cell for the day Percentage of snow is then calculated as percentage snow 100 count of snow observations count of land observations Cloud percentage of a CMG cell is calculated in the same way as the percentage of snow except that count of cloud observations is used The same calculation is used because only land cells same as those for snow calculation are included in the calculation Cloud percentage is stored in the Day CMG Cloud Obscured SDS The confidence index was developed to provide users with an estimate of confidence in the snow value reported for cell Confidence index values are stored in the Day Confidence Index SDS This index indicates how confident the algorithm is that the snow percentage in a cell is a good estimate based on data snow snow free land cloud other binned into the grid cell A high is indicative of cloudless conditions and good data values and that the snow percentage reported is a very good estimate A low Cl is indicative o
102. within a selected data range Data used to fill gaps in the swath Used for oceans Nominal grid cell resolution Decision point to process a cell as land or water Antarctica masked as perennial snow cover Key to meaning of data in the SDS HDF predefined attribute names I3 latitude longitude 0 100 255 254 0 05 deg 12 00000 Antarctica deliberately mapped as snow QA value set to 252 0 good quality 1 other quality 252 Antarctica mask 253 data not mapped 254 mask 255 fill Primary sources of snow errors in MOD10C 1 are the result of snow errors being propagated from the MOD10 L2 through the MOD10A1 product into the MOD10C1 product Snow errors are typically manifest as lower fractions 1 25 range of fractional snow in the map These snow errors are generally scattered around the globe but may be more frequent in temporal and spatial extent in some regions Pattern of the snow errors on any day may have an appearance related to cloud cover for the day if the source of the error is snow cloud confusion or cloud shadowed land A user may want to mask all or part of this range 1 25 of fractional snow from use depending on application and interpretation by the user Errors originating from causes described above are most obvious in temperate and subtropical climates in the summer months During the summer months the errors may be screened from use by various methods that remove the er
103. zed products from the sequence of products Therefore understanding the assimilation of accuracy and error between levels and through higher levels is necessary to make optimal use of the products Description of assimilated error and how it affects the accuracy of the product is included in each product section A user may want to study the preceding product s description to enhance their understanding of the product accuracy MODIS Terra and MODIS Aqua versions of the snow products are generated This user guide applies to products generated from both sensors but is written based primarily on the Terra products Bias to Terra is because the snow detection algorithm is based on use of near infrared data at 1 6 um primary key to snow detection is the characteristic of snow to have high visible reflectance and low reflectance in the near infrared MODIS band 6 MODIS band 6 1 6 um on Terra is fully functional however MODIS band 6 on Aqua is only about 30 functional 7096 of the band 6 detectors non functional That situation on Aqua caused a switch to band 7 2 1 uim for snow mapping in the swath level algorithm The bias to Terra is also because of the greater understanding of the MODIS Terra sensor pre launch algorithm development 1 of 80 longer data record of Terra and greater amount of testing the Terra algorithms in preparation for Collection 5 processing Discussion of reasons for the different bands and the effect on snow mapping a

MODIS Snow Products User Guide to Collection 5 George A. Riggs

Contents

Download Pdf Manuals

Related Search

Related Contents