Circa 2000 Landsat ETM+ mosaic of northern Canada

Metadata:

Identification_Information:

Citation:

Citation_Information:

Originator:

Government of Canada, Natural Resources Canada / Earth Sciences Sector / Canada Centre for Remote Sensing;

Publication_Date:  2005

Title: Circa 2000 Landsat ETM+ mosaic of northern Canada

Edition: 1

Geospatial_Data_Presentation_Form:  Digital, Raster Data, Remote-Sensing Data  

Publication_Information:

Publication_Place: Ottawa, Ontario, Canada

Publisher:

Government of Canada, Natural Resources Canada / Earth Sciences Sector / Canada Centre for Remote Sensing

Other_Citation_Details:

Online_Linkage: ftp:\ccrs.nrcan.gc.ca\AD\EMS\

Description:

Abstract:

This product resulted from a joint collaboration between the Climate Change and Groundwater Programs at the Earth Sciences Sector of Natural Resources Canada, and was supported by the Canadian Space Agency through the Government Related Initiatives Program (GRIP) proposal “EO-based view of our landmass in support of northern development”.

A mosaic of northern Canada above the tree line was produced from public domain Landsat ETM+ level L1G bands 3, 4 and 5 ortho-imagery available through the Centre for Topographic Information of Natural Resources Canada. Radiometric normalization was necessary due to different acquisition times among the 446 individual scenes in this mosaic. A highly automated radiometric normalization technique was performed to match scenes to a common radiometric reference consisting of a coarse (1km) apparent reflectance 2000 growing season composite image generated at CCRS from VGT-S01 data from SPOT’s VEGETATION sensor. Normalization involved resampling individual Landsat scenes to 1 km resolution using the VGT point spread function and generating band-specific normalization coefficients between Landsat and VGT using a robust regression technique called Thiel-Sen. This technique has been shown to produce optimal consistency among Landsat scenes in Olthof et al., 2005, and is highly automated allowing production of a mosaic of this size.

These data are available in geo-tiff format with 57173 columns and 32914 rows at 90 m resolution obtained by resampling 30 m Landsat data. More information on the processing chain used to generate these data are available in Olthof, Butson, Fernandes, Fraser, Latifovic and Orazietti (2005). Landsat ETM+ mosaic of northern Canada. Canadian Journal of Remote Sensing (accepted).

 

Purpose:

The purpose if this work was to generate a spatially consistent dataset of northern Canada with sufficient detail required by scientists and other users interested in studying processes in the Canadaian Arctic regions above the tree line.

Supplemental_Information:  Language of dataset: English

Time_Period_of_Content:   

Time_Period_Information:  

Range_of_Dates/Times:  July – August, 1999 - 2002

Beginning_Date:  July 14, 1999

Beginning_Time:

Ending_Date:  August 28, 2002

Ending_Time:

Currentness_Reference: Metadata

Status:

Progress:

Maintenance_and_Update_Frequency: As needed

Spatial_Domain:

Bounding_Coordinates:

West_Bounding_Coordinate: 158d35'02.88" W

East_Bounding_Coordinate: 27d52'24.52" W

North_Bounding_Coordinate: 65d24'05.43" N

South_Bounding_Coordinate: 43d28'32.05" N

Keywords:

Theme:

Theme_Keyword_Thesaurus: 

Theme_Keyword:

Earth Science > Northern Biosphere > Vegetation Composition/structure > Vegetation Cover

Theme_Keyword: Earth Science > Human Dimensions > Infrastructure

Theme_Keyword:

Earth Science > Land Surface > Land Cover > Land Use

Theme_Keyword: Earth Science > Climate Change

Theme_Keyword: Earth Science > North > Northern

Theme_Keyword: Earth Science > Land Surface > Landscape > Tree Line

Theme_Keyword: Earth Science > Land Surface > Topography > Landforms

Theme_Keyword: Earth Science > Cryosphere > Snow/ice

Theme_Keyword: Earth Science > Hydrosphere > Snow/ice

Theme:

Theme_Keyword_Thesaurus: userDefined

Theme_Keyword: Landsat

Theme_Keyword: TM

Theme_Keyword: ETM+

Theme_Keyword: Radiance

Theme_Keyword: Mosaic

Place:

Place_Keyword_Thesaurus:

Place_Keyword: National Region > Arctic

Place_Keyword: National Region > Northern

Place_Keyword: Global Regions > North America

Place_Keyword: Countries > Canada

Place_Keyword: Earth Science > Inner Atmosphere > Land Surface

Place:

Place_Keyword_Thesaurus: derived

Place_Keyword: Canada > Northwest Territories

Place_Keyword: Canada > Nunavut

Place_Keyword: Canada > Yukon

Place_Keyword: Canada > Ontario

Place_Keyword: Canada > Quebec

Place_Keyword: Canada > Labrador

Place:

Access_Constraints: None

Use_Constraints:

Her Majesty the Queen in Right of Canada, 1999. For more information, please contact the Copyright Officer at the following address: ESS Copyright Officer Business Development Office Earth Sciences Sector 615 Booth St., Ottawa, Ontario K1A 0E9 Sa Majesté la Reine du chef du Canada, 1999 Pour plus d’information, prier de communiquer avec le responsable du droit d’auteur Ă  l’adresse suivante : Responsable du droit d’auteur du SST Bureau d’expansion des affaires Secteur des sciences de la Terre 615, rue Booth Ottawa (Ontario) K1A 0E9

Point_of_Contact:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Dr. Ian Olthof

Contact_Organization:

Government of Canada, Natural Resources Canada / Earth Sciences Sector / Canada Centre for Remote Sensing

Contact_Position: Visiting Fellow Research Scientist

Contact_Address:

Address_Type: Mailing and physical address

Address: 588 Booth Street

City: Ottawa

State_or_Province: Ontario

Postal_Code: K1A 0Y7

Country: Canada

Contact_Voice_Telephone: +01-613- 947-1233

Contact_Facsimile_Telephone: +01-613- 947-1406

Contact_Electronic_Mail_Address: iolthof@CCRS.NRCan.gc.ca

Data_Quality_Information:

Lineage:

Source_Information:

Source_Citation:

Citation_Information:

Originator:

Publication_Date:

Geospatial_Data_Presentation_Form:

Type_of_Source_Media:

Source_Time_Period_of_Content:

Time_Period_Information:

Range_of_Dates/Times:

Beginning_Date:

Beginning_Time:

Ending_Date:

Ending_Time:

Source_Currentness_Reference:

Process_Step:

Process_Description:

 

9. DATA MANIPULATIONS

The normalization and compositing methods used to create this mosaic were designed specifically to maintain high consistency among 446 individual Landsat scenes acquired on different dates. To achieve optimal radiometric consistency, a reference dataset consisting of a peak of growing season composite from SPOT’s VEGETATION (VGT) satellite was used. The reference data are apparent at-surface reflectance bands 2, 3 and 4 from VGT, whose wavelengths and bandwidths almost identically match the red, Near Infrared (NIR) and Shortwave Infrared (SWIR) bands 3 (0.63-0.69um), 4 (0.75-0.90um) and 5 (1.55-1.75um) from Landsat. Daily data acquired at coarse (1 km) resolution and advanced compositing techniques provide consistent nation-wide clear sky observations at peak of growing season to serve as a radiometric reference for medium resolution Landsat data. The normalization technique exploits the best features of two data types by combining the high temporal frequency and large area consistency of coarse resolution data with the high spatial detail available from medium resolution sensors.

Step 1. Normalization function generation

Landsat scenes were resampled to 1-km resolution using the VGT point spread function and then randomly sampled to generate approximately 1200 1-km Landsat pixels per scene with spatially coincident pixels from the VGT imagery. These data were used to generate Landsat scene-wise normalization functions using Thiel-Sen regression separately for red, NIR and SWIR bands with the Landsat data as the regressor and VGT as the response variable. Once the set of normalization coefficients were finalized, scaling factors (Table 1) were applied to all normalization functions to minimize data compression or saturation of each band during normalization, as in Du et al. (2001). The scaled Thiel-Sen gain and offset coefficients were then applied to the 1-km Landsat imagery on a per-scene and per-band basis. The few scenes that did not normalize well were identified with a visual quality check of the mosaic and adjusted by applying coefficients from an adjacent scene acquired during the same orbit, or by adjusting the offset slightly to minimize the appearance of seam lines.

 

           

Table 1. Scaling factors

Red

3

NIR

1.5

SWIR

1.6

 

Step 2. Compositing

Landsat images were resampled to 90 m due to the prohibitive large size at 30 m resolution by averaging 30-m pixels on a 3 by 3 pixel basis, producing an effective spatial resolution of approximately 98m in x and 94m in y when the point spread function of the ETM+ sensor is accounted for, and a thus minimum mapping unit of 0.92 hectares. The normalized scenes were first mosaiced north to south along paths and east to west along rows. A second mosaic was generated by entering scenes in the opposite order, i.e. south to north along paths and west to east along rows. Each of these two mosaics represented one of two overlap areas between adjacent scenes. The best data in overlap regions from the two mosaics was selected using a maximum NDVI criterion to produce the final image data, as this criterion has been widely shown to preferentially select clear sky pixels over haze or cloud for vegetated surfaces. Masks were applied to the final mosaic to remove unwanted data such as sea ice in the far north.

Process_Date: 2004

Spatial_Data_Organization_Information:

Direct_Spatial_Reference_Method: Raster

Raster_Object_Information:

Raster_Object_Type: Pixel

Row_Count: 32914

Column_Count: 57173

Spatial_Reference_Information:

Horizontal_Coordinate_System_Definition:

Planar:

Map_Projection:

Map_Projection_Name: Lambert Conformal Conic

Lambert_Conformal_Conic:

Standard_Parallel: 1st Parallel 49d N, 2nd Parallel 77d N

Longitude_of_Central_Meridian: 95d W

Latitude_of_Projection_Origin: 49d N

False_Easting: 0

False_Northing: 0

Planar_Coordinate_Information:

Planar_Coordinate_Encoding_Method: Row and column

Coordinate_Representation:

Abscissa_Resolution:  90 m

Ordinate_Resolution: 90 m

Planar_Distance_Units: Meters, feet

Geodetic_Model:

Horizontal_Datum_Name: NAD83

Ellipsoid_Name: E008

Map corners (in metres):

Upper Left     -2410973.000 E         3104929.000 N

Upper Right     2734597.000 E         3104929.000 N

Upper Left     -2410973.000 E          142669.000 N

Lower Right     2734597.000 E          142669.000 N

Pixel Size        90 E      90 N

Map corners (in degrees) :

Upper Left  Corner  158d35'02.88" W  Lon   65d24'05.43" N  Lat

Upper Right Corner   27d52'24.52" W  Lon   62d50'14.75" N  Lat

Image Center             91d36'24.72" W  Lon   63d50'21.00" N  Lat

Lower Left Corner   126d13'02.48" W  Lon   44d53'14.88" N  Lat

Lower Right Corner    60d20'22.90" W  Lon   43d28'32.05" N  Lat

Entity_and_Attribute_Information:

Detailed_Description:

Entity_Type:

Entity_Type_Label: Pixel

Attribute:

Attribute_Label: Radiance

Attribute_Definition: Numeric code ranging from 0 to 255

Attribute_Definition_Source:

Attribute_Domain_Values:

Enumerated_Domain:

Enumerated_Domain_Value:

Distribution_Information:

Distributor:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Ian Olthof

Contact_Organization:

Government of Canada, Natural Resources Canada, Canada Centre for Remote Sensing

Contact_Position: Visiting Fellow Research Scientist

Contact_Address:

Address_Type: Mailing and physical address

Address: 588 Booth Street

City: Ottawa

State_or_Province: Ontario

Postal_Code: K1A 0Y7

Country: Canada

Contact_Voice_Telephone: +01-613- 947-1233

Contact_Facsimile_Telephone: +01-613- 947-1406

Contact_Electronic_Mail_Address: iolthof@CCRS.NRCan.gc.ca

Distribution_Information:

Distributor:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Bill Park

Contact_Organization: Canada Centre for Remote Sensing

Contact_Position: Physical Scientist

Contact_Address:

Address_Type: Mailing and physical address

Address: 588 Booth St.

City: Ottawa

State_or_Province: Ontario

Postal_Code: K1A 0Y7

Country: Canada

Contact_Voice_Telephone: +01-613-947-1371

Contact_Facsimile_Telephone: +01-613-947-1406

Contact_Electronic_Mail_Address: Bill.Park@CCRS.NRCan.gc.ca

Distribution_Information:

Distributor:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Ian Olthof

Contact_Organization:

Government of Canada, Natural Resources Canada, Canada Centre for Remote Sensing

Contact_Position: Visiting Fellow Research Scientist

Contact_Address:

Address_Type: Mailing and physical address

Address: 588 Booth Street

City: Ottawa

State_or_Province: Ontario

Postal_Code: K1A 0Y7

Country: Canada

Contact_Voice_Telephone: +01-613- 947-1233

Contact_Facsimile_Telephone: +01-613- 947-1406

Contact_Electronic_Mail_Address: iolthof@CCRS.NRCan.gc.ca

Distribution_Liability: None

Standard_Order_Process:

Digital_Form:

Digital_Transfer_Information:

Format_Name: TIFF

Digital_Transfer_Option:

Online_Option:

Computer_Contact_Information:

Network_Address:

Network_Resource_Name:

Offline_Option:

Offline_Media:

8mm Cartridge Tape CD-ROM

Recording_Format: CD-ROM

Fees: Free

Ordering_Instructions:

Users may place requests by letter, telephone, electronic mail, FAX, or personal visit.

Standard_Order_Process:

Non-digital_Form: Map

Fees: Free

Ordering_Instructions:

Users may place requests by letter, telephone, electronic mail, FAX, or personal visit.

Metadata_Reference_Information:

Metadata_Date: 200510

Metadata_Contact:

Contact_Information:

Contact_Person_Primary:

Contact_Person: Ian Olthof

Contact_Organization:

Government of Canada, Natural Resources Canada, Canada Centre for Remote Sensing

Contact_Position: Visiting Fellow Research Scientist

Contact_Address:

Address_Type: Mailing and physical address

Address: 588 Booth Street

City: Ottawa

State_or_Province: Ontario

Postal_Code: K1A 0Y7

Country: Canada

Contact_Voice_Telephone: +01-613- 947-1233

Contact_Facsimile_Telephone: +01-613- 947-1406

Contact_Electronic_Mail_Address: iolthof@CCRS.NRCan.gc.ca

Metadata_Standard_Name:

Metadata_Standard_Version:

Metadata_Time_Convention: Local Time