Ancillary Data
Several ancillary data sets are also available in EASE-Grid. These data sets were derived from external data sources and then gridded to the EASE-Grid. These data sets help to answer questions such as what percentage of the EASE-Grid pixels are a particular type of land cover classification, and what is the average elevation at each EASE-Grid pixel. Choose one of the following categories to learn more and to download files:
See the EASE-Grid Geolocation Tools Web page for grids that contain latitudes and longitudes of EASE-Grid pixels, along with C, IDL, and Fortran subroutines for converting coordinates.
Land Cover
IGBP Land Cover Classes
The Boston University MOD12Q1 V004 Land Cover Product (BU-MODIS) and Global Land Cover Characterization (GLCC) data sets in EASE-Grid were derived using the International Geosphere Biosphere Programme (IGBP) land cover classes. Table 1 lists the IGBP land cover classes.
| Class Number | Category |
|---|---|
| 1 |
Evergreen needleleaf forest
|
| 2 |
Evergreen broadleaf forest
|
| 3 |
Deciduous needleleaf forest
|
| 4 |
Deciduous broadleaf forest
|
| 5 |
Mixed forests
|
| 6 |
Closed shrublands
|
| 7 |
Open shrublands
|
| 8 |
Woody savannas
|
| 9 |
Savannas
|
| 10 |
Grasslands
|
| 11 |
Permanent Wetlands
|
| 12 |
Croplands
|
| 13 |
Urban and built-up
|
| 14 |
Cropland/natural vegetation mosaic
|
| 15 |
Snow and ice
|
| 16 |
Barren or sparsely vegetated
|
| 17 |
Water bodies
|
BU-MODIS Land Cover [top]
Table 2 lists the Boston University MOD12Q1 V004 Land Cover Product (BU-MODIS) data files in EASE-Grid. The IGBP land cover classes were used and are listed in Table 1. The original BU-MODIS file was the 1 km Interrupted Goode Homolosine Projection. Each tar file contains a series of compressed flat data files for one projection and resolution.
| Region | Resolution | Grid Size | File Name |
|---|---|---|---|
| Northern Hemisphere | 25 km | 721 x 721 | Nl.BU-MODIS.tar.gz |
| 12.5 km | 1441 x 1441 | Nh.BU-MODIS.tar.gz | |
| Southern Hemisphere | 25 km | 721 x 721 | Sl.BU-MODIS.tar.gz |
| 12.5 km | 1441 x 1441 | Sh.BU-MODIS.tar.gz | |
| Global | 25 km | 1383 x 586 | Ml.BU-MODIS.tar.gz |
| 12.5 km | 2766 x 1171 | Mh.BU-MODIS.tar.gz |
Compressed tar files can be extracted using the following command:
% gzip -c file.tar.gz | tar xvf -
Each flat data file named grid.gl_g.sds01.v4.??.bin has the proportion of land per pixel for a single class stored as percent. So, Nl.gl_g.sds01.v4.01.bin is a 721x721 by 1 byte file (stored by row) where each byte represents the percent of land in that cell which was characterized as IGBP class 1. Pixels which are off the map, or were undefined in the original have the value 255.
Processing Notes
The original BU-MODIS 1 km data were resampled to EASE-Grid by aggregating up to 25 km by drop-in-the-bucket averaging (no weighting factor). The floating point percent value was rounded to the nearest integer percent. There were three problems with the result:
- Water/missing ambiguity: For some reason, BU used 0 for water instead of the IGBP value of 17. Since this is a land product there were large areas in the oceans that were assigned 255 (fill) instead of 0 (water).
- Round-off error: the sum of the integer percent values at any pixel location ranged from 98 to 102, due to round-off error.
- Anomalous water in the Ross Ice Shelf: due to an unexplained artifact in the original 1 km data, about 10 resampled pixels in the middle of the Ross Ice Shelf were coded as partially open water, rather than the expected snow and ice category (IGBP class 15).
In order to produce a more consistent data set, NSIDC performed the following corrections:
- Water layer 17 was derived from BU-MODIS classes 0 through 16 as: water layer = 100 - (sum of nonwater layers, 1-16)/(sum of all layers, 0-16)
- The integer sum at each pixel was normalized to 100 by examining the original floating point values, and reversing what would have rounded up or down for those values with decimal parts closest to 0.5, for only enough classes to achieve a sum of 100. Since non-normalized sums ranged from 98 to 102, this adjusted the rounding in any given category by at most 3 percentage points.
- The misclassified open water in the Ross Ice Shelf was changed to be all category 15.
- The pixel containing the North Pole in the Northern Hemisphere was undefined in the original. It was set to 100 percent water.
To acknowledge these EASE-Grid BU-MODIS land cover data in a publication, use the following citation:
Knowles, K. 2004. EASE-Grid Land Cover Data Resampled from Boston University Version of Global 1 km Land Cover from MODIS 2001, Version 4. Boulder, Colorado USA: National Snow and Ice Data Center. Digital Media.
BU-MODIS Land-Ocean-Coast-Ice (LOCI) Masks
Corresponding Land-Ocean-Coast-Ice (LOCI) masks derived from the BU-MODIS land cover data are included in each BU-MODIS tar file. A .png version of each LOCI file is also included for quick illustration of the contents. LOCI files are named according to the following convention:
(region)_loci_land50_coast0km.(grid).bin
where(region)indicates spatial coverage (such as Northern Hemisphere, Southern Hemisphere, or Global) and (grid) indicates grid size. For example, Nl_loci_land50_coast0km.721x721.bin is the LOCI file for the Northern Hemisphere with a grid size of 721 columns by 721 rows.
LOCI files are in flat binary 1-byte format with the following Digital Number values:
0 = land
101 = ice
252 = coast
254 = off_the_map
255 = water
The LOCI masks are derived as follows: The IGBP classes 1-16 are summed. Any cell > 50% is labeled land. Any cell with >= 50% water (igbp 17) is labeled ocean. Any cell with >= 50% ice (igbp 15) is labeled ice. The LOCI masks available here do not include a coastline classification. We have produced versions of these masks with expanded coastlines, but they are not included here. Contact NSIDC User Services if you are interested.
To acknowledge these EASE-Grid LOCI masks in a publication, use the following citation:
EASE-Grid Land-Ocean-Coastline-Ice (LOCI) Mask, Derived from Boston University Version of Global 1 km Land Cover from MODIS 2001, Version 4, Courtesy of K. Knowles. Boulder, Colorado USA: National Snow and Ice Data Center. Digital Media.
Note: NSIDC first distributed the BU-MODIS land cover tar files in September 2006. In July 2009, an alert user notified us that the tar files were missing layer 10 (IGBP code for GRASSLANDS). We have corrected the omission. As of 31 July 2009, the tar files now include the missing layer 10. We apologize for any inconvenience.
GLCC Land Cover [top]
Table 3 lists the Global Land Cover Characterization (GLCC) data files in EASE-Grid. The IGBP land cover classes were used and are listed in Table 1.
| Region | Resolution | Grid size | File Name |
|---|---|---|---|
| Northern Hemisphere | 25 km | 721 x 721 | Nlglcc.tar |
| 12.5 km | 1441 x 1441 | Nhglcc.tar | |
| Southern Hemisphere | 25 km | 721 x 721 | Slglcc.tar |
| 12.5 km | 1441 x 1441 | Shglcc.tar | |
| Global | 25 km | 1383 x 586 | Mlglcc.tar |
| 12.5 km | 2766 x 1171 | Mhglcc.tar |
Each tar file contains a series of compressed flat data files for one projection and resolution. Each flat data file has the proportion of land per pixel for a single class stored as percent. For example, Nligbp01.dat is a 721 x 721 1-byte file (stored by row) where each byte represents the percent of land in that cell characterized as IGBP class 1. Pixels that are off the map or undefined in the original land cover map have a value of 255.
To extract a file, using water (class 17) from the 25 km Northern Hemisphere grid as an example, type the following command at a unix command prompt:
% tar xvf Nlglcc.tar
You will see the following output from this command:
x Nligbp01.dat.gz, 35579 bytes, 70 blocks
x Nligbp02.dat.gz, 22045 bytes, 44 blocks
x Nligbp03.dat.gz, 13565 bytes, 27 blocks
x Nligbp04.dat.gz, 26790 bytes, 53 blocks
x Nligbp05.dat.gz, 41744 bytes, 82 blocks
x Nligbp06.dat.gz, 24185 bytes, 48 blocks
x Nligbp07.dat.gz, 57880 bytes, 114 blocks
x Nligbp08.dat.gz, 50607 bytes, 99 blocks
x Nligbp09.dat.gz, 21303 bytes, 42 blocks
x Nligbp10.dat.gz, 50574 bytes, 99 blocks
x Nligbp11.dat.gz, 17230 bytes, 34 blocks
x Nligbp12.dat.gz, 62909 bytes, 123 blocks
x Nligbp13.dat.gz, 15851 bytes, 31 blocks
x Nligbp14.dat.gz, 63472 bytes, 124 blocks
x Nligbp15.dat.gz, 9178 bytes, 18 blocks
x Nligbp16.dat.gz, 43568 bytes, 86 blocks
x Nligbp17.dat.gz, 62126 bytes, 122 blocks
Then type:
% gzip -dv Nligbp17.dat.gz
You will see the following output from this command:
Nligbp17.dat.gz: 88.0% -- replaced with Nligbp17.dat
To acknowledge these EASE-Grid AVHRR land cover data in a publication, use the following citation:
Knowles, K. 2004. EASE-Grid Land Cover Data Resampled from AVHRR Global 1 km Land Cover, Version 2, March 1992 - April 1993. Boulder, Colorado USA: National Snow and Ice Data Center. Digital Media.
Elevations [top]
These are the Global Land One-km Base Elevation (GLOBE) data regridded to EASE-Grid. They consist of 2-byte integers in meters with values of -500 indicating no data. The .dat files have a big-endian, network byte order suitable for most Unix systems, including Sun and SGI. The .pc files have little-endian, host byte order suitable for Intel PC machines running Windows or Linux.
|
|
To acknowledge these EASE-Grid elevation data in a publication, use the following citation:
Knowles, K. 2001. EASE-Grid Elevation Data Resampled from the Global Land One-km Base Elevation (GLOBE) Project. Boulder, Colorado USA: National Snow and Ice Data Center. Digital media.
Pixel Area [top]
Data files are not provided because the EASE-Grids are based on equal-area map projections. Every pixel in a grid has the same area:
25 km grids (North, South, Global): 628.3827 sq km
12.5 km grids (North, South, Global): 157.0957 sq km
