I'm re-posting my question becouse i've experienced some problems with
my e-mail service.
It's a newby question so sorry for bothering.
After reading Goovaerts' Geostatistics for Natural Resources Evaluation
and Isaaks and Srivastava's Applied Geostatistics I still haven't found
a way to corectly choose among variogram models.
Is a spherical model linked closer to a particular type of spatial
process? Are continous and smooth processes better described by
sphercial or and exponential model? Choosing between a sphericl or an
exponetial model is just a matter of fitting to experimental variogram
or is there specific reason on how the process under analysis works?
As an ecologist I have to provide a "meaningful" explanation of why
using a specific kind of model to describe my experimental variogram of
an difference image of NDVI values derived from Landsat satellite.
Any suggestion will be helpful.
Thanks to all.
NiCoLa
***********************************************************************
*
* Dott. Nicola Zaccarelli
* Lab. di Ecologia del Paesaggio
* Dip. di Scienze e Tecnologie Biologiche ed Ambientali
* Universit� di Lecce
*
***********************************************************************
Prov.le Lecce-Monteroni
73100 - LECCE
Tel. (+39)- 0832-320.896
Fax: (+39)- 0832-320.626
e-mail: [EMAIL PROTECTED]
[EMAIL PROTECTED]
NiCoLa
___________________________________________
Dott. Nicola Zaccarelli
C.N.R.-I.I.A.
tel. 0832-320.886
tel.
0832-320.896��,����Ơz�-��b�� �ح�����,��l�x�r,j�z�����ǭ��^�ǫ��^jǫzg�m����,��.�f�ʇڟ+�y����)�{ ��2����-���N�����r��zǧu�ޙ�������ݢj.�)\���z,���r֧v�칻 �&�j( ��Z���Z0y��zwh��^��m�)ފ{azg��� n�r
�NM!
+����m�Ԟ˛���m�ު笶�h����R��h��h����b���'^u�a���
0j( ��Z��+
