## GEOSTATS: spatial correlation method

Expand Messages
• Hi all, I am very new to spatial statistics and I have the following problem: I have an Arc/INFO coverage with two roads: one is from the existing road network
Message 1 of 2 , Oct 31, 2000
• 0 Attachment
Hi all,

I am very new to spatial statistics and I have the following problem:
I have an Arc/INFO coverage with two roads: one is from the existing road
network and the other one is the same road modelled with a road modelling
software.
How can I assess the spatial correlation between these two representations

Thank you,

Ionut

Ionut Aron, MF Candidate
Forest Resources Management
Forest Sciences Centre
2045-2424 Main Mall
University of British Columbia
Telephone: (604) 822-4148

--
*To post a message to the list, send it to ai-geostats@....
*As a general service to list users, please remember to post a summary
of any useful responses to your questions.
*To unsubscribe, send email to majordomo@... with no subject and
"unsubscribe ai-geostats" in the message body.
DO NOT SEND Subscribe/Unsubscribe requests to the list!
• ************************************************************** This list has moved. Please see the end of this message for instructions. ... If your
Message 2 of 2 , Nov 1, 2000
• 0 Attachment
**************************************************************
This list has moved. Please see the end of this message for
instructions.
**************************************************************

> How can I assess the spatial correlation between
> these two representations
> of the same road ?

If your information has the same spatial co-ordinates
in both sets, you can produce a co-located cross
semi-variogram by simply calculating the covariance
for each lag:

(g_i - g_j)(f_i - f_j)

where g and f are the two variables, i and j are two
locations a specified distance apart. Average over all
pairs the same distance apart and divide by two. This
is exactly analogous to calculating the covariance
between the two at that specified distance:

(g_i - gbar)(f_j - fbar)

only you don't need the two means. If your samples are
not at the same locations you need the non-co-located
semi-variogram, referred to by some authors as a
"pseudo cross semi-variogram" and given as standard in
Noel Cressie's book. The basic form of this is

(g_i - f_j)^2

notation as above. This assumes that the two variables
have the same mean or have been standardised somehow.

Kriging is always the same. Watch out for Volume 2 of
Practical Geostatistics -- PG2001. Lots of
explanations there.

Isobel Clark
drisobelclark@...
http://uk.geocities.com/drisobelclark
Isobel Clark

____________________________________________________________
Do You Yahoo!?