reading a .tif file in R [closed]

Question

I am reading a .tif file in R and getting 4 warnings messages listed below. When I follow the instruction on the 4th message, the first 3 warnings still remain but the values read from file changes drastically at every pixel. Please help me in reading the data correctly from the .tif files. Sample File can be found on the link: ftp://ftp.ntsg.umt.edu/pub/MODIS/NTSG_Products/MOD16/MOD16A2_MONTHLY.MERRA_GMAO_1kmALB/GEOTIFF_0.05degree/

my code:

remove(list=ls()) 

library(tiff)

library(raster)

str_name<-'MOD16A2_ET_0.05deg_GEO_2008M01.tif' 

read_file<-readTIFF(str_name) 

Warning messages:

1: In readTIFF(str_name) :
  TIFFReadDirectory: Unknown field with tag 33550 (0x830e) encountered
2: In readTIFF(str_name) :
  TIFFReadDirectory: Unknown field with tag 33922 (0x8482) encountered
3: In readTIFF(str_name) :
  TIFFReadDirectory: Unknown field with tag 34735 (0x87af) encountered
4: In readTIFF(str_name) :
  tiff package currently only supports unsigned integer or float sample formats in direct mode, but the image contains signed integer format - it will be treated as unsigned (use native=TRUE or convert=TRUE to avoid this issue)

Please help me with this issue of reading tif files correctly. Thanks in advance.

Answer 1

Did you try simply the raster package raster function (or stack if multiple layered tif)? The raster package was made to deal with geo-referenced raster datasets:

library(raster)
str_name<-'MOD16A2_ET_0.05deg_GEO_2008M01.tif' 
imported_raster=raster(str_name)

The simple code above works and yields a raster object with the following properties:

class       : RasterLayer 
dimensions  : 2800, 7200, 20160000  (nrow, ncol, ncell)
resolution  : 0.05, 0.05  (x, y)
extent      : -180, 180, -60, 80  (xmin, xmax, ymin, ymax)
coord. ref. : +proj=longlat +datum=WGS84 +no_defs +ellps=WGS84 +towgs84=0,0,0 
data source : C:\Users\lfortini\Downloads\MOD16A2_ET_0.05deg_GEO_2000M01.tif 
names       : MOD16A2_ET_0.05deg_GEO_2000M01 
values      : -32768, 32767  (min, max)

Answer 2

Simply read the pixels as unsigned and convert them to signed:

 t = readTIFF("MOD16A2_ET_0.05deg_GEO_2008M01.tif", as.is=TRUE)
 t[t >= 32738L] = -65536L + t[t >= 32738L]

Looking at the image, you may also want to convert -32768 to NA as that seems to be the use in the file:

 t[t == -32768L] = NA

If you want to convert the integers to [-1,1] reals now, just do

 t = t / 32768

The first three warnings are just telling you that there are additional custom tags in the file.