Data preparation¶

Copying Landsat sample products to the Sandbox¶

Log on the Sandbox shell and run:

curl http://landsat.usgs.gov/documents/L5_30m19910616.tgz | tar xvfz -
curl http://landsat.usgs.gov/documents/L5_30m19950627.tgz | tar xvfz -
curl http://landsat.usgs.gov/documents/L7_30m19990817.tgz | tar xvfz -
curl http://landsat.usgs.gov/documents/L7_30m20090422.tgz | tar xvfz -

This will download and extract the files from the compressed archives.

Format conversion¶

Use GDAL to convert the several GeoTIFF files (one for each Landsat band) to ERDAS .img format by creating a GDAL Virtual Format [1]

On the Sandbox shell run the command:

gdalbuildvrt -separate myvrt L5043033_03319950627_B10.TIF \
  L5043033_03319950627_B20.TIF \
  L5043033_03319950627_B30.TIF \
  L5043033_03319950627_B40.TIF \
  L5043033_03319950627_B50.TIF \
  L5043033_03319950627_B60.TIF \
  L5043033_03319950627_B70.TIF

and finally:

gdal_translate -of HFA myvrt L5043033_03319950627.img

Repeat the two steps above with the other three Landsat products.

Copying the ERDAS .img products to S3 storage¶

On the Sandbox shell run the command:

s3cmd put *.img s3://<your_laboratory>-private/data/

List the uploaded files:

s3cmd ls s3://<your_laboratory>-private/data/

Registering the ERDAS .img products in the Sandbox local catalogue¶

To register the Landsat products converted to ERDAS .img format, you need to create a dataset series which is a container for the datasets.

Copy the contents of the file below into a file named series.rdf in your home.

Landsat series

For each Landsat product, generate a file containing the dataset metadata.

Finally go to the Sandbox catalogue Web Interface at the address http://<sandbox IP>/catalogue/search and click search, you will see the Landsat products!

Footnotes