Aligning large raster and SpatialPoints in R

0
votes

I have two datasets:

  1. A raster file containing per-pixel vegetation classifications across Australia.

The file is described to have projection of EPSG:3577 https://epsg.io/3577. All additional meta data, along with data itself can be found at http://www.agriculture.gov.au/abares/forestsaustralia/forest-data-maps-and-tools/spatial-data/forest-cover

  1. A csv file containing data a latitude column, a longitude column, and various other variables.

As this is a large dataset also, I will only provide the head of this:

firms <- structure(list(latitude = c(-20.5897, -22.4119, -21.132, -23.9083, 
-23.5908, -23.689), longitude = c(147.6421, 148.8484, 148.1897, 
147.2966, 150.1676, 150.0994), brightness = c(320.7, 316.7, 320.5, 
320.1, 315.8, 314.3), scan = c(1.6, 2.1, 1.8, 1.7, 2.7, 2.7), 
    track = c(1.3, 1.4, 1.3, 1.3, 1.6, 1.6), acq_date = c("2017-01-01", 
    "2017-01-01", "2017-01-01", "2017-01-01", "2017-01-01", "2017-01-01"
    ), acq_time = c(47L, 47L, 47L, 47L, 47L, 47L), satellite = c("Terra", 
    "Terra", "Terra", "Terra", "Terra", "Terra"), instrument = c("MODIS", 
    "MODIS", "MODIS", "MODIS", "MODIS", "MODIS"), confidence = c(34L, 
    26L, 36L, 53L, 33L, 22L), version = c(6.2, 6.2, 6.2, 6.2, 
    6.2, 6.2), bright_t31 = c(299.6, 295.3, 299.7, 296.6, 291.7, 
    289.3), frp = c(19, 25.8, 22.9, 17.6, 35.4, 30), daynight = c("D", 
    "D", "D", "D", "D", "D"), type = c(0L, 0L, 0L, 0L, 0L, 0L
    )), row.names = c(NA, -6L), class = c("data.table", "data.frame"
), .internal.selfref = <pointer: 0x10200e2e0>)

My goal is simple:

Plot the raster data and then plot the each row from the csv file as a point on top of the raster.

I am happy with plot(), levelplot(), or ggplot() although the latter seems silly as it takes too much time to convert this massive raster to a data.frame.

So far my code looks like this:

library(rgdal)
library(raster)
library(rasterVis)
library(sp)
  1. Load in the raster file 
test <- raster('w001000.adf')
  1. Then create coordinates for the data.frame and make it a SpatialPoints format
coordinates(firms) <- ~longitude + latitude
firms_pts <- SpatialPoints(coords=coordinates(firms), proj4string = CRS("+proj=aea +lat_1=-18 +lat_2=-36 +lat_0=0 +lon_0=132 +x_0=0 +y_0=0 +ellps=GRS80 +towgs84=0,0,0,0,0,0,0 +units=m +no_defs"))
  1. Transform the points data into the same crs as the raster
firms_lyr <- spTransform(firms_pts, crs(test))

When I try to plot the data at this point, I cannot see any of the points.

  1. So I use the same extent in both datasets and now then plot roughly on top of each other.
extent(test) <- extent(firms)
  1. plot
plot(test)
points(firms_lyr, pch=16, cex=0.1)

NOW... annoyingly, the data sets are not properly aligned spatially this

Clearly this is something to do with the projection.

I am pretty sure I am doing something wrong in steps 3 and/or 4

1
rplotspatialraster

1 Answers

1
votes

Is this working? I am not sure if the points are the ones in the far east or not.

data

firms <- structure(list(latitude = c(-20.5897, -22.4119, -21.132, -23.9083,  -23.5908, -23.689), 
                        longitude = c(147.6421, 148.8484, 148.1897, 147.2966, 150.1676, 150.0994)), 
                   .Names = c("longitude", "latitude"), 
                   class = "data.frame", row.names = c(NA, -6L))

xy <- firms[,c(2,1)]
firms_pts <- SpatialPointsDataFrame(coords = xy, data = firms,
                       proj4string = CRS("+proj=longlat +datum=WGS84 +ellps=WGS84 +towgs84=0,0,0"))

Shapefile reprojection using adf

library(raster)
setwd(paste(dirname(rstudioapi::getActiveDocumentContext()$path), "/aus_for18", sep=""))
test <- raster("aus_for18.adf")
firms_lyr <- spTransform(firms_pts, crs(test))

Plotting

plot(test)
points(firms_lyr, pch=16, cex=1) #Queensland points

enter image description here