RGISTools

Handling multiplatform satellite images.

Table of contents

• The package
• Installation
• Credentials for downloading satellite images
• Copyright and license

The package

This package enables you downloading, customizing, and processing time series of satellite images from Landsat, MODIS and Sentinel in a standardized way. Some functions download and convert automatically the platform-specific file formats into GTiff, so they can be loaded in R. The customization functions support tile mosaicking, cropping, cloud masking and deriving new variables of interest, such as the NDVI, EVI, etc. Tile mosaicking is required when the region of interest extends over several tiles, so they can be combined into a single image. Cropping involves removing the pixels outside the region of interest, making any analysis more computationally and memory efficient. Cloud masking eliminates cloud reflectance that would otherwise be erroneously attributed to land surface features. Cloud removal and (measurement or processing) errors trigger data gaps and outliers, decreasing the quality and quantity of measurements. Hence, the package includes a set of function for filling and smoothing the satellite imagery. The combination of functions in RGISTools results in a stack of satellite images ready-to-use. Due to the wide variety of procedures and sources of information being handled in RGISTools, the functions are divided into 7 categories, which are identified by the first 3 characters of the function names;

1. mod identifies Modis Terra and Aqua satellite functions.
2. sen identifies Sentinel functions.
3. ls7 identifies Landsat 7 functions.
4. ls8 identifies Landsat 8 functions.
5. ls identifies both Landsat 7 and 8 functions.
6. gen identifies function for being used in any of the three platforms.
7. var identifies function for deriving variables in any of the three platforms.

Below, there is a list of the most important functions grouped by platform, and listed in operational order. These functions include searching, previewing, downloading, mosaicking, deriving new variables, compositing, cloud masking and filling/smoothing satellite imagery.

I. Landsat functions

The Landsat program is currently releasing imagery captured by two satellites; the Landsat-7 and Lansat-8. Both satellites are treated separately in coding terms due to discrepancies in their spectral coverages and data formats. To download Landsat imagery with the following functions, a USGS's EarthExplorer account is required. Please, register here.

Landsat-7

• ls7LoadMetadata Loads the Landsat-7 metadata file.
• ls7Search Seeks a time series of Landsat-7 images.
• lsPreview Previews Landsat satellite images.
• lsDownSearch Downloads a time series of Landsat images.
• lsMosaic Mosaics Landsat images.
• ls7FolderToVar Computes new variables from Landsat-7 multispectral images.
• lsCloudMask Creates cloud masks for Landsat images.
• genSaveTSRData Saves a time series of images.

Landsat-8

• ls8LoadMetadata Loads the Landsat-7 metadata file.
• ls8Search Seeks a time series of Landsat-7 images.
• lsPreview Previews Landsat satellite images.
• lsDownSearch Downloads a time series of Landsat images.
• lsMosaic Mosaics Landsat images.
• ls8FolderToVar Computes new variables from Landsat-7 multispectral images.
• lsCloudMask Creates cloud masks for Landsat images.
• genSaveTSRData Saves a time series of images.

II. MODIS functions

Functions in RGISTools download all land products from Terra and Aqua satellites, but the processing focuses on the multispectral images. Be aware that an EarthData account is required to use NASA's web service so, please, register here.

• modSearch Seeks a time series of MODIS images.
• modPreview Previews MODIS satellite images.
• modDownSearch Downloads a time series of MODIS images.
• modMosaic Mosaics MODIS images.
• modFolderToVar Computes new variables from MODIS multispectral images.
• modCloudMask Creates cloud masks for MODIS images.
• genSaveTSRData Saves a time series of images.

III. Sentinel functions

Sentinel archives provide a wide variety of products based on a 5-satellite constellation. The functions to download Sentinel images can cope with any product available in ESA's SciHub web service. However, image processing is focused on Sentinel-2 multispectal images. SciHub credentials are required to download Sentinel imagery and can be obtained here.

• senSearch Seeks a time series of Sentinel images.
• senPreview Previews Sentinel images.
• senDownSearch Downloads a time series of Sentinel images.
• senMosaic Mosaics Sentinel images.
• senCloudMask Creates cloud masks for Sentinel images.
• senFolderToVar Computes new variables from Sentinel-2 multispectral images.
• genSaveTSRData Saves a time series of images.

IV. Important general functions

In addition to functions above, the package provides some general functions for a better data handling:

• genCompositions Creates image compositions from a time series of satellite images.
• genSmoothingIMA Fills the gaps and smooths outliers in a time series of satellite images.
• genSmoothingCovIMA Fills the gaps and smooths outliers in a time series of satellite images using covariates.
• genPlotGIS Plots satellite images with a proper GIS format.
• genGetDates Gets the capturing date of an image from the name of a raster layer.

V. Remote sensing variables

New variables can be derived from multispectral images. The most common variables in the scientific literature are pre-programmed in RGISTools. They can be identified by the prefix "var".

• varEVI Calculates the enhanced vegetation index (EVI).
• varMSAVI2 Calculates the modified soil-adjusted vegetation index (MSAVI2).
• varNBR Calculates the normalized burn ratio (NBR).
• varNBR2 Calculates the normalized burn ratio 2 (NBR2).
• varNDMI Calculates the normalized difference moisture index (NDMI).
• varNDVI Calculates the normalized difference vegetation index (NDVI).
• varNDWI Calculates the normalized difference water index (NDWI).
• varRGB Calculates an RGB image from 3 spectral bands.
• varSAVI Calculates the soil-adjusted vegetation index (SAVI).

Installation

Install from CRAN

# Install RGISTools package
install.packages("RGISTools")

# load RGISTools library
library(RGISTools)

Install from GitHub

# Install devtools package from cran repository
install.packages("devtools")

# load devtools library
library(devtools)

# Install RGISTools from GitHub repositoy
install_github("spatialstatisticsupna/RGISTools")

Dependencies for linux

The package depends on some R packages that in Linux requires the installation of some libraries before the installation in R. Here you have the command to install all the applications from repository for Debian/Ubuntu and RedHat/Fedora.

Debian/Ubuntu

sudo apt update
sudo apt install r-cran-rcpp gdal-bin libgdal-dev libproj-dev libssl libssl-dev xml2 libxml2-dev libmagick++-dev

RedHat/Fedora

sudo dnf install gdal gdal_devel proj_devel xml2 libxml2_devel libcurl_devel openssl_devel ImageMagick-c++_devel

Credentials for downloading satellite images

Modis

Credentials EarthData

Landsat

Credentials EarthData

Sentinel

Credentials SciHub

Copyright and license

Licensed under the GPL-3 License. Full license here.