subset gridded data as geotiff and plot"

library(rmarkdown)
options(continue=" ")
options(width=60)
library(knitr)
library(geoknife)
knitr::opts_chunk$set(eval=FALSE)

Service is no longer working. This vignette will not be built.

geoknife has a number of output formats, but a new one that was added in version 1.1.0 of the package (after the initial release to CRAN) is the ability to output results as a zip file that contains a series of geotiffs for each timestep requested. This vignette is a brief introduction on how to do this using a few additional packages that are in the Suggests field in the description.

get the data

see the geoknife getting started vignette for more details, but for the purposes of this vignette, we need to get some data first and then plot it up.
First things first, load up the geoknife package

library(geoknife)

For this example, I am going to use a dataset that is hosted by the USGS, and includes an estimate of actual evapotranspiration. There is a lot of data here, but I am just going to pluck a subset in time (one timepoint for the month of July of one year) and a fairly large spatial subset. I am going to use the OUTPUT_TYPE='geotiff' argument that was added in version 1.1.0 of geoknife.

knife <- webprocess(algorithm = list('OPeNDAP Subset' = "gov.usgs.cida.gdp.wps.algorithm.FeatureCoverageOPeNDAPIntersectionAlgorithm"))

fabric <- webdata(url='https://cida.usgs.gov/thredds/dodsC/ssebopeta/monthly', 
                  variable="et", #Monthly evapotranspiration
                  times=c('2014-07-15','2014-07-15'))

stencil <- simplegeom(data.frame('point1' = c(-76,49), 'point2' = c(-93,40)))
job <- geoknife(stencil, fabric, knife, wait = TRUE, OUTPUT_TYPE = "geotiff")

now that the job is complete (because wait=TRUE was used), we can download the result and unzip it:

dest <- file.path(tempdir(), 'et_data.zip')

file <- download(job, destination = dest, overwrite = TRUE)

unzip(file, exdir=file.path(tempdir(),'et'))

tiff.dir <- file.path(tempdir(),'et')

plot the data

using the rasterVis package, read this in and create a raster object:

library(rasterVis)
library(raster)

et <- raster(file.path(tiff.dir , dir(tiff.dir)))
NAvalue(et) <- -1

Now, use ggplot2 and ggmap to plot this as a map:

library(ggmap)
library(ggplot2)

gplot(et, maxpixels = 5e5) + 
  geom_tile(aes(fill = value), alpha=1) +
  scale_fill_gradientn("Actual ET (mm)", colours=rev(heat.colors(5))) +
  coord_equal() + 
  theme_classic() + 
  theme(axis.line = element_blank(), axis.text.x = element_blank(), axis.text.y = element_blank(),
        axis.ticks = element_blank(), axis.title.x = element_blank(), axis.title.y = element_blank()) +
  scale_y_continuous(limits=c(et@extent@ymin, et@extent@ymax)) + 
  scale_x_continuous(limits=c(et@extent@xmin, et@extent@xmax))


Try the geoknife package in your browser

Any scripts or data that you put into this service are public.

geoknife documentation built on July 9, 2023, 5:31 p.m.