How to read *.adf files into R?

Question

I want to load .adf files into R. The data is from this page: http://www.fao.org/geonetwork/srv/en/metadata.show?id=14057

I tried the following code that I found after some research in the internet. The problem is, that in the class RasterLayer I get negativ values that shouldn't be there. I don't know why this happens, so hopefully someone can help me!?

Code:

library(rgdal)
library(RColorBrewer)
dpath<- path...

x <- new("GDALReadOnlyDataset", dpath)
getDriver(x)
getDriverLongName(getDriver(x))
xx<-asSGDF_GROD(x)
r <- raster(xx)

The output for 'r' is:

r class : RasterLayer dimensions : 2160, 4320, 9331200 (nrow, ncol, ncell) resolution : 0.08333333, 0.08333333 (x, y) extent : -180, 180, -90, 90 (xmin, xmax, ymin, ymax) coord. ref. : +proj=longlat +ellps=WGS84 +towgs84=0,0,0,0,0,0,0 +no_defs data source : in memory names : band1 values : -997, 16 (min, max)

The '16' in the values refers to the 16 classes of length of growing period. But I wonder where those '-997' come from. Maybe something wrong with the coord. ref?

Here is also a data summary of 'xx':

Data summary: Min. 1st Qu. Median Mean 3rd Qu. Max. NA's -997 3 5 -9 8 16 7123158

And if we look at the data in xx more closely:

table(xx$band1)

-997 1 2 3 4 5 6 7 8 9 10 11 12 31711 429643 83011 166674 207228 270161 240958 183342 118608 98795 88473 73743 56022 13 14 15 16 30104 45521 52216 31832

There is really just this '-997' thing in it. I think the NAs are ozeans, so is there something wrong with the data loading or do I just don't understand the data?

Answer 1

You are almost right:

NODATA is set to -32768 for oceans. Additionally, -997 is set for great lakes that are not excluded by the coastline.

Since the pixel content (growing period) makes no sense on lakes, you can safely treat -997 as NODATA too.