[ai-geostats] Jacknifing in GSLIB and stubborn data points
- Dear all,
I've been trying to use the 'jacknifing' function in GSLIB, with Ordinary Kriging, with two data sets. The first is a set of hard elevation points (numbering about 283) over an area of a river valley, taken by GPS. The second is a set of 323 (approx) elevation points which are often very close to the hard elevation points, and whose values are unsampled and to which I'm trying to assign predicted elevations through the kriging process. I've done the variogram modelling etc, but when it comes to applying the method it all works very well for predicting most of the 323 points, however, I am always left with a little clustered group of about 7 points which always give a value of -999 (which means 'unestimated'). I've tried tweaking and fiddling around a bit with the variogram model parameters, but all to no avail. This stubborn select little clustered group refuse to be 'predicted' and remain 'unestimated'. I can't see that their pattern is at all special, or that they are in anyway 'different' to other data which has kriged perfectly well.
I am wondering if anyone might know or might have a hunch as to what is happening and what I should do to get these points to 'give in' and be estimated.
Many thanks, in anticipation.
This email may contain privileged/confidential information.
It is intended solely for the person to whom it is addressed.
If you are not the intended recipient you may not copy, deliver or disclose the content of this message to anyone.
In such case please destroy/delete the message immediately and notify the sender by reply email.
Opinions, conclusions and other information in this message that does not relate to the official business of Sevenoaks District Council shall be understood as neither given nor endorsed by the Council.
Visit the Council at WWW.SEVENOAKS.GOV.UK
This message has been checked for all known viruses by Star Internet delivered through the MessageLabs Virus Control Centre.