This repository contains R scripts that can be used and edited for creating pixel-based image composites for Landsat TM/ETM+ data. Other Landsat sensors (MSS/OLI) have not yet been implemented. The scripts were adapted from a workflow described in "Van doninck J. and Tuomisto H. (2018) A Landsat composite covering all Amazonia for applications in ecology and conservation. Remote Sensing in Ecology and Conservation, 4(3), 197-210".
0. Wrapper script
The script OGIIR_0_Wrapper_Example.R provides a simple example on how the scripts can be used for downloading and processing Landsat TM/ETM+ data. Parameters set in this example script can be changed by the users to define their own spatial and temporal area of interest.
1. Query and download Landsat L1 data from USGS EarthExplorer
The script OGIIR_1_browseEE.R provides an interface to the USGS/EROS Inventory Service Documentation. Use of this services requires an ERS user profile authorization for Machine to Machine access, which must be requested separately (firstname.lastname@example.org).
2. Process L1 data to directionally normalized surface reflectance
The script OGIIR_2_srTopoBRDF.R uses Ledaps and other USGS-EROS scripts to convert Level-1 data to surface reflectance, to mask clouds and cloud cover, and to derive Landsat viewing angles. The script further provides options for normalization of topographic effects, if a digital elevation model is provided by the user, and for BRDF normalization.
USGS-EROS ESPA product formatter (https://github.com/USGS-EROS/espa-product-formatter) and ESPA Ledaps (https://github.com/USGS-EROS/espa-surface-reflectance/tree/master/ledaps) must be installed to be able to run these scripts.
3. Image projection and compositing
The scripts OGIIR_3_composite.R provides options for projecting a set of Landsat images to a common coordinate system and extent, and for performing pixel-based image compositing using a number of compositing criteria (maximum NDVI, band-wise median, multidimensional median).
When used, the following citing should be mentioned: "We made use of geospatial data/instructions/computing resources provided by the Open Geospatial Information Infrastructure for Research (oGIIR, urn:nbn:fi:research-infras-2016072513) funded by the Academy of Finland.