## GEOSTATS: Geostatistical Software Conventions

Expand Messages
• Dear AI- GEOSTATS readers, Geostatistical programs differ in the way they parameterize semivariogram models and in the way they consider direction. As I tried
Message 1 of 3 , Aug 27, 1998

Geostatistical programs differ in the way they parameterize semivariogram
models and in the way they consider direction. As I tried to understand
with the AI-GEOSTATS readership. If these issues are important to your
then I will update these notes. Gregoire has agreed to put them somewhere
on the AI-GEOSTATS web page. Thanks.

Sincerely,
Tom Mueller

muelle26@...
http://www.msu.edu/user/muelle26/

A summary of Some Geostatistical Software Conventions
by Tom Mueller

Exponential Model

(1) semivariogram = c[1-exp(-h/a)] - SAS, Surfer, GSLIB(1992)
(2) semivariogram = c[1-exp(-3h/a)] - GS+, Variowin, GSLIB(1998)

If you use an exponential model to fit experimental semivariograms with
Variowin, you must divide a (vw) by three to krig with Surfer.

Gaussian Model

(1) semivariogram = c[1-exp((-h/a)^2)] - SAS, Surfer, GSLIB (1992)
(2) semivariogram = c[1-exp(-3(h/a)^2)] -Variowin, GS+
(3) semivariogram = c[1-exp((-3h/a)^2)] - GSLIB (1998)

If you use a Gaussian model to fit experimental semivariograms with
Variowin, you must divide a (vw) by the square root of three to krig with
Surfer or multiply a (vw) by the square root of three to krig with GSLIB
(1998).

Notes:
h = the separation vector

a = the range, practical range, effective range or range parameter

GS+ uses exponential model #1 and Gaussian model #1 computationally (they
refer to a as the range parameter) but you enter and read the range (not the
range parameter) as defined as the distance at which the semivariogram
reaches 95% of the sill. In other words, GS+ defines the range as 3 * a for
the exponential model and sqrt (3) * a for the Gaussian model. (This is why
above I have GS+ listed as using the second forms of the models above)
Variowin refers to this as the practical range and GSLIB refers this to this
as the effective range (but the ways in which Variowin and GSLIB calculate
them are different).

Surfer and Variowin consider the East direction to be 0 degrees and
direction to increases counterclockwise.

GS+, GSLIB (1992, 1998), and SAS considers due North to be 0 degrees and
direction to increase clockwise.

--
*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!
• A few additions/corrections to Tom s post about software conventions. I ve been using S+Spatialstats, and have added it to the appropriate lists below, with
Message 2 of 3 , Aug 27, 1998
I've been using S+Spatialstats, and have added it to the appropriate lists
below, with the convention that "a" is the value typed in as the "range"
in the variogram menu. Also, I think the negative sign is in the wrong
place in two of the Gaussian models (see below).

-Jennifer

Jennifer Dungan | MS 242-4
CSU - Monterey Bay | NASA Ames Research Center
Tel: 650-604-3618 FAX: 650-604-4680 | Moffett Field, CA 94035-1000
email: jdungan@... | USA
URL: http://geo.arc.nasa.gov |

On Thu, 27 Aug 1998, Tom Mueller wrote:

> A summary of Some Geostatistical Software Conventions
> by Tom Mueller
>
> Exponential Model
>
> (1) semivariogram=c[1-exp(-h/a)] - SAS, Surfer, GSLIB(1992)
> (2) semivariogram=c[1-exp(-3h/a)] - GS+,Variowin,GSLIB(1998),S+

> If you use an exponential model to fit experimental semivariograms with
> Variowin, you must divide a (vw) by three to krig with Surfer.
>
> Gaussian Model
>
> (1) semivariogram = c[1-exp((-h/a)^2)] - SAS, Surfer, GSLIB(1992), S+

should be c[1-exp(-(h/a)^2)]

> (2) semivariogram = c[1-exp(-3(h/a)^2)] -Variowin, GS+
> (3) semivariogram = c[1-exp((-3h/a)^2)] - GSLIB (1998)

should be c[1-exp(-(3h/a)^2)]
>

--
*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!
• Dear AI-GEOSTATS readers: Thanks to everyone who responded. I ve updated the original message and Gregoire has posted it on the AI-GEOSTATS web site,
Message 3 of 3 , Aug 30, 1998

Thanks to everyone who responded. I've updated the original message and
Gregoire has posted it on the AI-GEOSTATS web site, specifically located at
http://curie.ei.jrc.it/software/soft_convention.htm. As you will note, some
new software has been added (GEOEAS and S+) but there are many programs that
have not been included (XGAM, ESRI software, etc.). So please let us know
and we will keep this list up to date.

Thanks,
Tom Mueller
muelle26@...

--
*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!
Your message has been successfully submitted and would be delivered to recipients shortly.