1 - Download Data

Eli Holmes

2021-08-31

Source: vignettes/1_download_data.Rmd

Download data from an ERDDAP server

This will download data from the CoastWatch ERDDAP server. Alter the URL to download from a different ERDDAP server or data set.

Use rerddap

require(devtools)
devtools::install_github("ropensci/rerddap")
devtools::install_github("rmendels/rerddapXtracto")

Specify where to save the downloaded data. Data from a griddap() call can be verbose, so you’ll probably want a folder for it.

fil_dir <- file.path(here::here(), "inst", "extdata")
if(!dir.exists(fil_dir)) dir.create(fil_dir)

We will download SST data from CoastWatch. Here is the data access page for that dataset.

lats <- c(40.375, 50.375)
lons <- c(-141.875, -120.875)
df_info <- rerddap::info("ncdcOisst21Agg_LonPM180")
df <- rerddap::griddap("ncdcOisst21Agg_LonPM180", latitude = lats, longitude = lons, time = c("2021-06-19", "2021-06-19"), fields = "sst")$data

df is a data frame with lat and lon.

head(df)
##                   time    lat      lon zlev   sst
## 1 2021-06-19T12:00:00Z 40.375 -141.875    0 17.10
## 2 2021-06-19T12:00:00Z 40.375 -141.625    0 17.25
## 3 2021-06-19T12:00:00Z 40.375 -141.375    0 17.45
## 4 2021-06-19T12:00:00Z 40.375 -141.125    0 17.62
## 5 2021-06-19T12:00:00Z 40.375 -140.875    0 17.68
## 6 2021-06-19T12:00:00Z 40.375 -140.625    0 17.61

There is no projection information for these data, but the meta data just says that it is on a uniform lat-lon grid. That tells us that at minimum it is "+proj=longlat". It is important to include this information so that we can combine this with other map data later.

Turn this into a raster

We can then turn this matrix into a raster using the raster package and the rasterFromXYZ() function.

df2 <- data.frame(x=df$lon, y=df$lat, z=df$sst)
ras <- raster::rasterFromXYZ(df2, crs = "+proj=longlat")

Plot the raster

We can do this with the raster package.

## Loading required package: sp
plot(ras)

We can use the cmocean package and use its thermal palette.

plot(ras, col = cmocean::cmocean("thermal")(100))

We can plot with ggplot2 also.

## Loading required package: ggplot2
# Plot
gg <- ggplot(df) +
  geom_raster(aes(lon, lat, fill = sst)) +
  scale_fill_gradient2(midpoint = mean(df$sst, na.rm = TRUE),
                       low = "blue",
                       mid = "white",
                       high = "red") +
  labs(x = NULL,
       y = NULL,
       fill = "Celcius",
       title = "Sea Surface Temperature (SST)")
gg

It’s a bit ugly. We might futz with the look and use the cmocean palette.

gg + theme_bw() +
  scale_x_continuous(limits = lons, expand = c(-0.01, -0.01)) +
  scale_y_continuous(limits = lats, expand = c(-0.01, -0.01)) +
  cmocean::scale_fill_cmocean(alpha=1) + theme_bw()
## Scale for 'fill' is already present. Adding another scale for 'fill', which
## will replace the existing scale.
## Warning: Removed 248 rows containing missing values (geom_raster).

The default raster plot is a bit deformed since it is long-lat on the x and y axis. We can see what it would look like in a different projection.

newcrs <- "+proj=wintri +lon_0=-125 +lat_1=46 +x_0=0 +y_0=0 +datum=WGS84 +units=m +no_defs"
ras_proj<- projectRaster(ras, crs=newcrs, over=TRUE)
plot(ras_proj)
title(stringr::str_sub(newcrs, 1, 12))

Save

We’ll save this to use later. You’ll want to change the data_dir to wherever you are saving your data. I am saving the data in different forms since in the examples to follow, different forms will be used.

data_dir <- file.path(here::here(), "data", "sample_raster.rda")
sample_raster <- list(df=df, raster=ras, lats=lats, lons=lons)
save(sample_raster, file=data_dir)