R wrapper for the SkyWatch API
R
Switch branches/tags
Nothing to show
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Failed to load latest commit information.
R
examples
man
.Rbuildignore
.gitignore
DESCRIPTION
NAMESPACE
README.Rmd
README.md
SkyWatchr-manual.pdf

README.md

SkyWatchr

R wrapper for the SkyWatch API developed by SkyWatch.

The SkyWatch API aims to provide a singular point to discover and access the world’s remote sensing datasets. It allows users to search satellite imagery and climate/atmospheric datasets by wavelength (band), cloud cover, resolution, location, date, etc. Available datasets include ACOS, AIRS, CAI, FTS-SWIR, Landsat-8, MOPITT, OCO2, Sentinel-2 and TES.

An API key for free access to the SkyWatch API can be requested in this link.

A video tutorial explaining how to install and use the SkyWatchr package can be found in this post in my blog.

Installation

The stable version (recommended) of the SkyWatchr package can be installed from CRAN by executing the following command line in an R console:

install.packages('SkyWatchr')

Alternatively, the development version can be installed from this GitHub repository. You'll need to install first the devtools package:

install.packages('devtools')

Then install the SkyWatchr package from this repository:

devtools::install_github('amsantac/SkyWatchr')

When installing from GitHub, users may need to install some ancillary packages before installing the SkyWatchr package:

install.packages(c('checkmate', 'htmlwidgets', 'knitr'))

Usage

First load the SkyWatchr package:

library(SkyWatchr)

Store in an object your API key:

api_key <- "your_personal_alphanumeric_api_key"

For convenience, users can set the global option SkyWatchr.apikey (recommended) once per session via:

options(SkyWatchr.apikey = api_key)

Then try some queries with the querySW function:

# One of the world largest landfills "Olusosun Dump" in Nigeria
querySW(time_period = "2015-8", longitude_latitude = "3.367358,6.566358,3.387358,6.586358")

If the SkyWatchr.apikey option is not set, then provide the API key:

# One of the world largest landfills "Olusosun Dump" in Nigeria
querySW(api_key, time_period = "2015-8", longitude_latitude = "3.367358,6.566358,3.387358,6.586358")

See below additional sample queries:

# An Imperial Oil refinery in Canada
res <- querySW(time_period = 2015, longitude_latitude = "-82.4169987,42.954811", data_level = 3)
View(res)

# The smoggiest city on Earth Ahvaz, Iran
querySW(time_period = "2015-06", longitude_latitude = "48.676074,31.321119", data_level = 3)

# Ahvaz, Iran in September, 2015
querySW(time_period = "2015-9", longitude_latitude = "48.676074,31.321119", data_level = 1, 
        wavelength_band = "red,green,blue")

# Other examples
querySW(time_period = "2009-12-25", longitude_latitude = "-71.1043443253471,-42.3150676015829", 
        data_level = 2)
querySW(time_period = "2009-12-25", longitude_latitude = "-71.1043443253471,-42.3150676015829")

querySW(time_period = "2016-07-11,2016-07-12", 
        longitude_latitude = "-71.1043443253471,-42.3150676015829,71.1043443253471,-42.3150676015829,71.1043443253471,42.3150676015829,-71.1043443253471,42.3150676015829,-71.1043443253471,-42.3150676015829",
        instrument_satellite = "Landsat-8", data_level = 1, max_resolution = 30, 
        max_cloudcover = 100, wavelength_band = "Blue")

# Queries can also be performed using objects of class Spatial (as defined by the sp package) 
# projected in geographic coordinates
ex1 <- data.frame(x = -71.1043443253, y = -42.3150676016, data = "point")
coordinates(ex1) <- ~ x + y
class(ex1)
querySW(time_period = "2015-8", longitude_latitude = ex1)

Call examples above retrieve objects of class data.frame. Output can also be printed as html by setting the output argument as html (see an html output example HERE):

# Get output as html
querySW(time_period = "2015-8", longitude_latitude = "3.367358,6.566358,3.387358,6.586358", 
        output = "html")
querySW(time_period = "2016-07-11,2016-07-12", 
        longitude_latitude = "-71.1043443253471,-42.3150676015829,71.1043443253471,-42.3150676015829,71.1043443253471,42.3150676015829,-71.1043443253471,42.3150676015829,-71.1043443253471,-42.3150676015829",
        instrument_satellite = "Landsat-8", data_level = 1, max_resolution = 30, 
        max_cloudcover = 100, wavelength_band = "Blue", output = "html")

Files can be downloaded using the downloadSW function (or clicking the links in the html output as shown above):

res <- querySW(time_period = "2015-06", longitude_latitude = "48.676074,31.321119", 
               data_level = 3)
View(res)

# Download all files
downloadSW(res)  # Warning: This will download a huge amount of data (> 50 Gb)

# Download a subset
downloadSW(res[c(1,3), ])

# Use an expresion to subset files to be downloaded
downloadSW(res, source == "MOPITT" & size_kb < 2400)

Download links expire 1 hour after being generated.

If needed, the boundary box enclosing the extent of a given dataset retrieved in the results data.frame can be generated using the getPolygon function:

res <- querySW(time_period = "2015-8", longitude_latitude = "-71.1043443253471,-42.3150676015829")
sppolygon <- getPolygon(res, index = 55)   # index refers to corresponding row in the data.frame

# To visualize it use spplot or packages such as mapview
library(mapview)
mapView(sppolygon)

For help on the functions implemented in the SkyWatchr package see the reference manual or run the following command lines in the R console:

?SkyWatchr
?'SkyWatchr-package'