|
|
MOD11_L2 LST Product
This product is generated using the MODIS sensor radiance data product
(MOD021KM), the geolocation product (MOD03), the cloud mask product (MOD35_L2),
the quarterly landcover (MOD12Q1), and snow product (MOD10_L2). The output
file contains SDSs of LST, quality assurance (QA), error in LST, emissivities
in bands 31 and 32, viewing zenith angle and time, latitude and longitude
(each set of latitude and longitude for every 5 scan lines
and 5 pixels), local attributes, and global attributes. This LST product
is generated by the generalized split-window LST algorithm (Wan and Dozier,
1996). For complete global coverage a MOD11_L2 LST product would
be generated for all swaths acquired in daytime and nighttime on the Earth
including the polar regions.
The algorithm and data product contents for MOD11_L2 are described
in the following sections.
Algorithm Description
A brief sketch of the LST algorithm for MOD11_L2 is described here for
the purpose of aiding the user in understanding and interpreting the data
product.
The LST retrieval in a MODIS swath is constrained to pixels that:
1. have nominal Level 1B radiance data,
2. are in clear-sky conditions at a
99% confidence defined in MOD35,
3. are on land or inland water.
In the V4 LST processing, LST retrieval is made for
lake and river pixels at clear-sky
conditions with a 66% and higher confidence defined
in cloud-mask MOD35 and for other land pixels in clear-sky at a 99% confidence,
in order to improve the consistency between the spatial LST distributions
over lakes and their surrounding lands.
Data inputs to the LST algorithm are listed in
Table
2.
Clouds are masked with the MODIS Cloud Mask data product (MOD35_L2).
Masking of oceans is done with the 1 km resolution land/water mask,
contained in the MODIS geolocation product (MOD03).
Table 2. MODIS data product inputs to the MODIS
LST algorithm for the MOD11_L2 product.
| ESDT |
Long Name |
Data Used |
| MOD021KM |
MODIS Level 1B Calibrated and
Geolocated Radiances |
EV_1KM_Emissive for MODIS bands:
31 (11.03 µm)
32 (12.02 µm) |
| MOD03 |
MODIS Geolocation |
Land/Water Mask
Height
Sensor Zenith Angles
Solar Zenith Angles
Latitude
Longitude
EV start time |
| MOD35_L2 |
MODIS Cloud Mask |
Cloud_Mask
Latitude (every 5 lines)
Longitude (every 5 pixels) |
| MOD07_L2 |
MODIS Atmospheric Profile |
Surface_Temperature
Water_Vapor |
| MOD12Q1 |
Land Cover |
Land_Cover_Type_1 |
| MOD10_L2 |
MODIS Snow Cover |
Snow Cover |
Emissivities in bands 31 and 32 are estimated by the classification-based
emissivity method (Snyder and Wan, 1998) according to land cover types
in the pixel determined by the input data in quarterly Land Cover (MOD12Q1)
and daily Snow Cover (MOD10_L2). In the first year of the MODIS LST production,
the landcover product used in the LST algorithm is the IGBP type land cover
product produced by the University of Maryland Landcover group based on
AVHRR data in the early years.
Since June 2001, the land-cover product generated from MODIS data
is used in the MODIS LST processing.
A large uncertainty may exist in such
estimated emissivities in semi-arid and arid areas. So the quality of the
MOD11_L2 product may be poor in these areas. Users are advised to use caution
applying the MOD11_L2 LST data to project-applications, especially in semi-arid
and arid areas.
Because band 22 is used in the 4-11 micron test to determine the cloudyness
of a pixel in the MODIS cloudmask algorithm, the noisy fourth channel in
band 22 produced quite a lot of (cloud) strips in the cloudmask product based
on the old A-side MODIS data (prior to October 30, 2000).
To avoid the strips caused by the noisy channels, the cloudmask in all
fourth channels of the scan cubes (one scan cub contains ten channels in
each band) is refined with the adjacent pixels in the third and fifth channels.
If both the adjacent pixels in the third and fifth channels are clear-sky
pixels at a 99% confidence (66% or higher for inland water pixels in V4),
the pixel in the fourth channel will be treated
as clear-sky pixel for the LST retrieval.
Scientific Data Sets (SDS)
The MODIS L2 LST product contains nine scientific data sets (SDSs): LST,
QC, Error_LST, Emis_31, Emis_32, View_angle, View_time, Latitude, and Longitude.
The first seven DSDs are for 1km pixels. The last two DSDs are coarse resolution
(5 km) latitude and longitude data. Each set of them correspond to a center
pixel of a 5 km by 5 km block of pixels in the LST SDS. A mapping relationship
of geolocation data to the first seven DSDs is specified in the global
attribute StructMetadata.0. The mapping relationship was created
by the HDF-EOS SDPTK toolkit during production. Geolocation data is mapped
to the first seven DSDs data with an offset = 2 and increment = 5. The
first element (0,0) in the geolocation SDSs corresponds to element (2,2)
in LST SDS, then increments by 5 in the cross-track or along-track direction
to map geolocation data to the LST SDS element.
Details are shown in
Table 3.
Table 3. The SDSs in the MOD11_L2 product.
SDS
Name |
Long
Name |
Number
Type |
Unit |
Valid
Range |
Fill
Value |
scale
factor |
add
offset |
| LST |
Land-surface
temperature |
uint16 |
K |
7500-65535 |
0 |
0.02 |
0.0 |
| QC |
Quality control for
LST and emissivity |
uint16 |
none |
0-65535 |
0 |
NA |
NA |
| Error_LST |
Land-surface
temperature error |
uint8 |
K |
1-255 |
0 |
0.04 |
0. |
| Emis_31 |
Band 31 emissivity |
uint8 |
none |
1-255 |
0 |
0.002 |
0.49 |
| Emis_32 |
Band 32 emissivity |
uint8 |
none |
1-255 |
0 |
0.002 |
0.49 |
| View_angle |
zenith angle of MODIS
viewing at the pixel |
uint8 |
deg |
0-180 |
0 |
0.5 |
0 |
| View_time |
Time* of Land-surface
Temperature observation
(* as local solar time) |
uint8 |
hrs |
0-240 |
0 |
0.1 |
0 |
| Latitude |
Latitude of every 5 scan
lines and 5 pixels |
float32 |
degree |
-90.0
to 90.0 |
-999.9 |
NA |
NA |
| Longitude |
Longitude of every 5 scan
lines and 5 pixels |
float32 |
degree |
-180.0
to 180.0 |
-999.9 |
NA |
NA |
Note that the Error_LST value is only an estimated value. It is quite
conservative in real clear-sky conditions. However, the effect of cloud
contaminations is not considered in the error estimation.
Local Attributes
Archived with the "LST" SDS are local attributes including the coefficients
of the calibration which converts the SDS value to real LST value in K.
HDF predefined local attributes (Table
4) describe characteristics of the data.
Table 4. HDF-predefined local attributes for SDS
LST in the MOD11_L2 product.
Attribute
Name |
Reserved
Label(s) |
Definition |
Sample Value |
| Label |
long_name |
Long Name of the SDS |
Land-surface Temperature |
| Unit |
units |
SI units of the data, if any |
K |
| Format |
Number Type |
How the data are stored |
uint16 (16-bit unsigned integer) |
| Range |
valid_range |
Max and min values within a
selected data range |
7500-65535 * |
| Fill Value |
_FillValue |
Data used to fill gaps in the
swath |
0 |
| Calibration |
scale_factor
add_offset
scale_factor_err
add_offset_err
calibrated_nt |
scaling factor
add offset
scaling factor error
add offset error
calibrated nt |
0.02
0.
0.
0.
0. |
* The number 65535 in uint16 may be shown as -1s in 16-bit integer by
some software toolkits,for example, by ncdump in the HDF toolkit.
The effective calibration formula for the "LST" SDS is
LST = the SDS data in uint16 *
0.02, giving a value in the range of 150-1310.7K.
Global Attributes
There are three global ECS attributes, i.e., CoreMetadata.0, ArchiveMetadata.0,
and StructMetadata.0, in the MOD11_L2 data product. Contents of
these global attributes were determined and written during generation of
the product and are used in archiving and populating the EOSDIS database
to support user services. They are stored as very long character strings
in parameter value language (PVL) format. Descriptions of the global attributes
are given here to assist the user in understanding them.
CoreMetadata.0 is the global attribute in which information compiled
about the product during product generation is archived and is used to
populate the EOSDIS database to support user services. The content of the
global attributes with sample values and comment of definition are listed
in Table 5, Table 6, and
Table
7, respectively. The user wanting detailed explanations of the
global attributes and related information should query the EOSDIS
related
web sites.
Table 5. Listing of objects in the global attribute
CoreMetadata.0
in MOD11_L2.
| Object Name |
Sample Value |
Comment |
| ShortName |
"MOD11_L2" |
ESDT name of product |
| VersionID |
1 |
ECS Version |
| ReprocessingActual |
"processed once" |
|
| ReprocessingPlanned |
"further update is anticipated" |
Expect that products will be
reprocessed one or more times. |
| LocalGranuleID |
"MOD11_L2.A2000207.1915.002.2000243053331.hdf" |
|
| DayNightFlag |
"Day" |
Day , Night or Both. |
| ProductionDateTime |
"2000-08-30T05:33:31.000Z" |
|
| LocalVersionID |
"SCF V2.2.16" |
Version of algorithm delivered
from the SCF. |
| PGEVersion |
"2.2.16" |
Version of production generation
executable. |
| InputPointer |
"MOD03.A2000207.1915.002.2000212171250.hdf","..." |
Location of input files in the
production system. |
| RangeBeginningDate |
"2000-07-25" |
Beginning and ending
times of the first and last scan line in the swath. |
| RangeBeginningTime |
"19:15:00.000000" |
| RangeEndingDate |
"2000-07-25" |
| RangeEndingTime |
"19:20:00.000000" |
| ExclusionGRingFlag |
"N" |
Geographic bounds
of swath coverage. |
| GRingPointLatitude |
[50.334011, 45.870819, 28.863239,
32.337044] |
| GRingPointLongitude |
[ -134.358658, -103.376228,
-112.166939, -136.159561] |
| GRingPointSequenceNo |
[1,2,3,4] |
| OrbitNumber |
3210 |
|
| EquatorCrossingLongitude |
-131.022286 |
|
| EquatorCrossingDate |
"2000-07-25" |
|
| EquatorCrossingTime |
"19:28:37.410935" |
|
| ParameterName |
"MODIS LST" |
|
| AutomaticQualityFlag |
"Passed" |
Result of automated checks during
the run of the algorithm that screen for significant amounts of anomalous
data. |
| AutomaticQuality |
"Passed if algorithm ran within
bounds of execution constraints. Suspect if bounds of execution constraints
violated. Failed if PGE failed." |
Explanation of result of automated
QA checks made during execution. |
| OperationalQualityFlag |
"Passed" |
Set by production system. |
| OperationalQualityFlagExplanation |
"Nominal Production" |
Explanation of Operational Flag |
| ScienceQualityFlag |
"Suspect" |
Set by LST investigator after
post-production investigation |
| SciencelQualityFlagExplanation |
"Early product assessment is
on going. Users are advised to use caution applying these data to
project-applications, especially in semi-arid and arid areas. Input
product are still being refined." |
Explanation of Science Flag |
| QAPercentMissingData |
0
(note that the value is incorrect in MOD11_L2 files
generated by earlier than PGEVERSION 2.2.14) |
0-100 |
| QAPercentCloudCover |
48
(for all pixels not VeryHighConfidentClear in MOD35_L2) |
0-100 |
| AncillaryInputPointer |
"MOD03.A2000207.1915.002.2000212171250.hdf" |
Location of geolocation input
product in production system. |
| AncillaryInputType |
"Geolocation" |
Type of ancillary data referenced
by pointer. |
| AssociatedSensorShortName |
"CCD" |
|
| AssociatedPlatformShortName |
"AM-1" |
|
| AssociatedInstrumentShortName |
"MODIS" |
|
| Product Specific
Attributes (PSA) |
| QAPercentGoodQuality |
34 |
Summary quality
assurance statistic for data product. Range is from 0-100. |
| QAPercentOtherQuality |
7 |
| QAPercentNotProducedCloud |
12 |
| QAPercentNotProducedOther |
47 |
| GranuleNumber |
233 |
Unique granule identifier |
| QAFractionGoodQuality |
0.3440894 |
Summary fraction
of the LST product. Range is from 0.0 to 1.0. |
| QAFractionOtherQuality |
0.0742003 |
| QAFractionNotProducedCloud |
0.1152801 |
| QAFractionNotProducedOther |
0.4664303 |
The four QAFraction PSAs are specially useful to granules in ocean regions
where only a small number of island pixels exist. Because the total
number of land and coastal pixels is highly variable in granules covering
both land and ocean, the values of QAPercent and QAFraction PSAs are calculated
on the base of the total number of all pixels in a granule. Therefore,
we can always calculate how many pixels with LST in good quality and other
quality from these PSA values, even for island pixels.
The ECS global attribute ArchiveMetadata.0 contains information
relevant to production of the data product. It also contains an alternate
bounding of geographic coverage of the swath. These data may be useful
in determining what version of the algorithm was used to generate the product.
Contents are described in Table 6.
Table 6. Listing of objects in the global attribute
ArchiveMetadata.0
in MOD11_L2.
| Object Name |
Typical Value |
Comment |
| EastBoundingCoordinate |
-103.376228 |
Extent of swath coverage, in latitude
and longitude. |
| WestBoundingCoordinate |
-136.159563 |
| NorthBoundingCoordinate |
50.334012 |
| SouthBoundingCoordinate |
28.863239 |
| AlgorithmPackageAcceptanceDate |
"1997-11-01" |
Algorithm Descriptors |
| AlgorithmPackageMaturityCode |
"pre-launch" |
| AlgorithmPackageName |
"MOD_PR11_V2" |
| AlgorithmPackageVersion |
"Version 2" |
| InstrumentName |
"Moderate-Resolution Imaging SpectroRadiometer" |
|
| AssociatedPlatformShortName |
"AM-1" |
|
| ProcessingDateTime |
"2000-08-30T05:33:31.000Z" |
|
| LongName |
"MODIS/Terra Land Surface Temperature/Emissivity
5-Min L2 Swath 1km" |
|
| Processing Center |
"GSFC" |
|
| SPSOParameters |
"2484 and 3323" |
|
| LocalInputGranuleID |
"MOD021KM.A2000207.1915..." |
input L1B HDF file. |
The StructMetadata.0 global attribute is used by the HDF-EOS
toolkit to specify the mapping relationships between the geolocation data
and the LST 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, or other software packages may be used to geolocate the data.
Table 7. Listing of objects in the global attribute
StructMetadata.0
in MOD11_L2.
| Object |
Definition |
| DIMENSION_1 |
10*nscans (along_swath_lines_1km) |
| DIMENSION_2 |
Max_EV_frames (Cross_swath_pixels_1km) |
| DIMENSION_3 |
2*nscans (Coarse_swath_lines_5km) |
| DIMENSION_4 |
Max_EV_frames/5 (Coarse_swath_pixels_5km) |
| DIMENSIONMAP_1 |
GeoDimension=2*nscans
DataDimension=10*nscans
Offset=2 Increment=5 |
| DIMENSIONMAP_2 |
GeoDimension=Max_EV_frames/5
DataDimension=Max_EV_frames
Offset=2 Increment=5 |
| GEOFIELD_1 |
GeoFieldName=Latitude |
| GEOFIELD_2 |
GeoFieldName=Longitude |
| DATAFIELD_1 |
DataFieldName=LST |
| DATAFIELD_2 |
DataFieldName=QC |
| DATAFIELD_3 |
DataFieldName=Error_LST |
| DATAFIELD_4 |
DataFieldName=Emis_31 |
| DATAFIELD_5 |
DataFieldName=Emis_32 |
| DATAFIELD_6 |
DataFieldName=View_angle |
| DATAFIELD_7 |
DataFieldName=View_time |
| DATAFIELD_8 |
DataFieldName=Latitude |
| DATAFIELD_9 |
DataFieldName=Longitude |
Quality Assurance
Indicators of quality are given in metadata objects in the CoreMetadata.0
global attribute QA and in a quality control (QC) SDS, generated during
production, or in post-product scientific and quality checks of the data
product. QA metadata objects in the CoreMetadata.0 global attribute
are the AutomaticQualityFlag and the ScienceQualityFlag and their corresponding
explanations. The AutomaticQualityFlag is set according to rules based
on data conditions encountered during a run of the LST algorithm. Setting
of this QA flag is fully automated. The rules used to set it are liberal;
nearly all of the data or intermediate calculations would have to be anomalous
for it to be set to "Failed". Typically, it will be set to "Passed" or
"Suspect". "Suspect" means that some bounds of execution constraints are
violated and that further analysis should be done to determine the source.
The ScienceQualityFlag is set post production either after an automated
QA program is run on the data product or after the data product is inspected
by a qualified LST investigator. Content and explanation of this flag are
dynamic so it should always be examined if present. A sampling of
products will be inspected. Sampling may be random, in support of field
campaigns, or event driven.
The QC SDS in the data product provides additional information on algorithm
results for each pixel. The QC SDS unsigned 16-bit data are stored as bit flags
in the SDS. This QC information can be extracted by reading the bits in
the 16-bit unsigned integer. The purpose of the QC SDS is to give the user
information on algorithm results for each pixel that can be viewed in a
spatial context. The QC information tells if algorithm results were nominal,
abnormal, or if other defined conditions were encountered for a pixel.
The QC information should be used to help determine the usefulness of the
LST data for a user's needs. The bit flags in the QC SDS are listed
in Table 8 .
Table 8. Bit flags defined in the QC SDS in the
MOD11_L2 product. Note that bit 0 is the least significant bit.
| bits |
Long Name |
Key |
| 1 & 0 |
Mandatory QA flags |
00=Pixel produced, good quality,
not necessary to examine
more detailed QA
01=Pixel produced, unreliable or unquantifiable quality,
recommend examination of more detailed QA
10=Pixel not produced due to cloud effects
11=Pixel not produced primarily due to reasons other
than cloud
(such as ocean pixel, poor input data) |
| 3 & 2 |
Data quality flag |
00=good
01=missing pixel
10=fairly calibrated
11=poorly calibrated, LST processing skipped |
| 5 & 4 |
Cloud flag |
00=cloud free pixel
01=pixel only with thin cirrus
10=fraction of sub-pixel clouds<= 2/16
11=TBD (in V3)
or
LST affected by nearby clouds (in V4) |
| 7 & 6 |
LST model number |
00=generalized split-window
method
01=day/night method
10=high LST w/o atmospheric & emis corrections
11=cirrus effects corrected |
| 9 & 8 |
LST quality flag |
00=no multi-method comparison
01=multi-method comparison done
10=fair consistency
11=good consistency |
| 11 & 10 |
Emissivity flag |
00=inferred from land cover
type
01=MODIS retrieved
10=TBD
11=default value used |
| 13 & 12 |
Emis quality flag |
00=emis quality not checked
01=emis quality checked with
land cover type
10=emis quality checked with NDVI
11=emis view-angle dependence checked |
| 15 & 14 |
Emis error flag |
00=error in emis_31 emis_32
<= 0.01
01=error in emis_31 emis_32 <= 0.02
10=error in emis_31 emis_32 <= 0.04
11=error in emis_31 emis_32 > 0.04 |
It should be noted that fillvalue 0 listed for the SDS QC
in Table 3 is valid for the bit flags only when
a fillvalue 0 is present in the SDS LST pixels (so the 01-00 bits in
the QC pxels have a value of 10 or 11). A value of 0 in the QC
bit flags means good data quality, cloud free, or small error in
emis_31 and emis_32, and etc, if a pixel has a valid LST value.
We do not discriminate fillvalue 0 from valid value 0 for all bit flags
in the QC in order to minimize the data volume.
Users should read SDSs LST and QC
at the same time in order to properly interpret their values in an easy way.
|