Workflow design

Data

You will use the Landsat 5&7 sample products made available by the USGS at http://landsat.usgs.gov/product_samples.php

The list of products used as test data is:

  • Landsat 5 TM

    Acquired June 16, 1991: L5 TM 30-meter thermal - .tgz (175 MB)

    Acquired June 27, 1995: L5 TM 30-meter thermal - .tgz (171 MB)

  • Landsat 7 ETM+

    Acquired August 17, 1999: L7 ETM+ 30-meter thermal - .tgz (316 MB)

    Acquired April 22, 2009: L7 ETM+ 30-meter thermal - .tgz (274 MB)

Software and COTS

GDAL and GDAL Python

You will use GDAL [1] to:

  • pre-process each of the Landsat sample products to convert from a multi GeoTIFF files to a single ERDAS .img product

and GDAL Python libraries [2] to:

  • implement a Python module to calculate the NDVI GeoTIFF

Workflow design

The application’s data pipeline activities can be defined as follows:

Use the Python NDVI package to apply the band arithmetic expression to calculate the NDVI to all Landsat products passed as references to the Sandbox catalogue.

!define DIAG_NAME Workflow example !include includes/skins.iuml skinparam backgroundColor #FFFFFF skinparam componentStyle uml2 start while (check stdin?) is (line) :Stage-in data; :Apply Python NDVI; :Stage-out ndvi_result; :Register ndvi_result in Sandbox catalogue; endwhile (empty) stop

This translates into a very simple workflow containing a single processing step: py-ndvi

The simple workflow can be represented as:

!define DIAG_NAME Workflow example !include includes/skins.iuml skinparam backgroundColor #FFFFFF skinparam componentStyle uml2 start :node_ndvi; stop

The node_ndvi is described in details in Processing nodes design

[1]GDAL Geospatial Data Abstraction Library
[2]GDAL Geospatial Data Abstraction Library Python package