Dear list members, I would like to ask your help to solve the following urgent problem. I am trying to use the option of block kriging in GSTAT, with an irregular block defined as the entire area under study. I would like to estimate a mean value and a variance from this area. I have used the following command file:
# gstat command file, Win32/Cygwin version 2.3.7 (12 July 2002) # data(ri1904): 'xyz1904.txt', x=1,y=2,v=3, average; variogram(ri1904): 74.3884sph(35780.3); set fit=0; area: 'blank50aw', mv=1; variances(ri1904): 'ri1904.var'; predictions(ri1904): 'ri1904.pre'; set output='ri1904.out'; The area file is an ASCII Surfer grid file, with grid cells 50 by 50 meter, with missing values set to 1 and other values set to 5. The area is about 25 km by 25 km, but irregularly shaped. The data set is a set of X and Y coordinates with Z (elevation) data of an estuary. It contains up to 4300 points. When running GSTAT, data are summed up with a sample mean of -2.309 and standard deviation of 5.076 (before kriging is applied). After the kriging procedure, the following output file is given: #x (map) y (map) pred(ri1904) var(ri1904) 332500 422500 0.110527 0.657318 Although the value of 0.65 m2 is smaller than the mean variance I have calculated for point kriging (0.89 m2), the values are much larger than anticipated, but I am not sure why... I would greatly appreciate any help in these matters. Kind regards, Daphne __________________________________________________ Dr Daphne van der Wal Netherlands Institute of Ecology (NIOO-KNAW)
