Loading ...
Sorry, an error occurred while loading the content.

AI-GEOSTATS: Summary of "Software and teaching"

Expand Messages
  • Francisco Jesús Moral García
    Hi all colleagues, This mail is to summarize all (numerous) answers to my questions (I hope everyone is included; forgive any omission). It has been very
    Message 1 of 1 , May 17, 2001
      Hi all colleagues,

      This mail is to summarize all (numerous) answers to my questions (I hope everyone is included; forgive any omission). It has been very helpful.

      Any further suggestion will be appreciated yet



      Marco Alfaro wrote,

      Habría que saber más acerca de tu problema. Si es bidimensional ó
      Si es bidimensional, creo que Gslib no aporta más.
      Tampoco sé si quieres aplicar geoestadística lineal (terminando en el
      krigeado) ó no lineal (simulaciones, krigeado disyuntivo, etc..). A qué
      variables quieres aplicar?


      Yo publiqué, en 1975 dos libros en España, yo no sé si aún quedan

      a) Introducción práctica a la Geoestadística.
      b) Geoestadística Operativa.

      Ambos fueron publicados por la Escuela de Ingenieros de Minas de Madrid.
      Rios Rosas 21, Madrid 3.


      Marco Alfaro


      Digby Millikan wrote,

      Hi Fransico,
      I am not familiar with individual papers on kriging, but can recommend many texts.

      "Practical Geostatistics 2000" Isobel Clark - The most popular book used at universities for a students
      textbook, which includes software.

      "Geostatistical Ore Reserve Estimation" M.David published by Elsevier. I found the easiest and best to
      understand as an introduction and application of point and block kriging, together with

      "Mining Geostatistics" A. Journel and Ch. Huijbregts enables the reader to program any kriging type and
      provides many examples of applications. These two early books, the first texts on geostatistics provide
      an excellent foundation and core of a geostatistical library. Though the later may be used selectively as
      it provides many invaluable sections such as variogram studies, example fortran point and block kriging
      programs, but also has sections containing some very advanced mathematics.

      Following this are a number of books available which provide you with insight as to what kriging is, and
      how it works.

      The first in this category is "An Introduction to Applied Geostatistics" M.Isaaks and R.Srivastava, which
      is based on a topography dataset used thoughout the text.

      "Basic Linear Geostatistics" M.Armstrong. This third generation up to date text of geostatistics covers
      the more basic kriging theory and brings together all of the important practical ways geostatistics can
      be used and applied in an easy to read format.

      "Geostatistics for Natural Resource Estimation" Pierre Goovaerts, Oxford press. Used as the text for a
      Graduate Diploma in Geostatistics at Michigan University (taught by Pierre) and as PhD entry text at
      Stanford University. Excellent text with many insights and pointers on applied geostatistics but a bit
      more advanced.

      Regarding software it may be usefull to determine what is contained in the Practical Geostatistics 2000
      package. I've never used GSLIB, very popular but I've heard it contains in cases understanding as to
      what input values are, as it is a compilation of software developed at Stanford, not designed for commercial
      use. It may also require a Fortran compiler. Maybe Compaq sells these.
      Also available is Isatis from Geovariances, France.

      Best Regards Digby Millikan.
      Geolite Mining Systems


      Gregoire Dubois wrote,

      Hello Francisco,

      I hope the following document (pdf format) will help.

      Best regards



      Frank Breen wrote,


      Surfer is a very powerful learning tool. The later versions also allow you to work directly with variograms which is how I learned geostats. GSLIB is also very good, but I found it to be rather advanced and cumbersome, unless you use it all the time. Another program I like is Sage2000 which will do 2D and 3D variograms. Also, check out Practical Geostatistics, it is a great book for self instruction and is modular and quite readable.

      Frank Breen

      Abedini wrote,

      Dear Francisco

      I am teaching a course in Geostatistics and my students found it helpful
      to work with gslib. FAQ of this mailing list might be a good start for
      teaching purposes.



      Carme Hervada wrote,

      Apreciado Francisco Jesús:
      llevamos ya cuatro (o cinco, no me acuerdo) años enseñando geoestadística a los ingenieros técnicos de minas. Durante este tiempo, hemos encontrado que el método gráfico mejor para personas técnicas, que no necesariamente saben programar y que están habituadas a utilizar el windows es el que tú usas.

      Ahora bien, para poder hacer investigación este método es muy limitado, muy encorsetado. En cambio, si utilizas los códigos fuente del GSLIB ( que están en Fortran) y los modificas y compilas según tus necesidades, tienes una mayor versatilidad, con lo que se puede hacer investigación mejor.

      Por lo que se refiere a artículos puedes entrar en la página web del ai-geostats ( que ahora no sé exactamente cuál es pero seguro que gregoire Dubois te lo dice) y allí encontrarás. De todos modos, hay algún libro, como el último de Ricardo Olea ( me parece que es Geostatistics for engineers and earth scientists) que no está de menos echarle un vistazo.

      Te sirve de algo?


      Paulo J. Ribeiro Jr. wote,

      Hola Francisco

      I believe there is no direct answer for your questions.
      The choice of geostatistical software (and there is plenty of it
      nowadays) might depends on the usage (teaching or research), methods
      required, approaches for geostatistics, public domain or comercial etc etc

      The AI-geostats web site has a large list of available software divided in
      categories which can help you

      For example, gslib offers a very nice set of Fortran/C code.
      Some available software are just a interface to these.
      However it is worth emphasise that GSLIB will have their approach to
      geostatistics. In other packages you might find different techniques
      because they take different views. For example, model-based methods
      (likelihood based and/or Bayesian inference) are the main motivation for
      geoR, a package I've been working on. This package, on its turn, will have
      several limitations.

      I would them use AI-list to select the "candidates", given your


      Wiliam Thayer wrote,

      I have worked with Gstat and GSLIB. GSLIB code is free but the Windows front
      end is not. As far as software goes, the price of GSLIB is very reasonable.
      Gstat is freeware and with it you can perform many types of geostatistical
      analyses. I am developing a Visual Basic/Map Objects interface that calls the
      Windows executable version of Gstat and have been satisfied with its
      capabilities to date. At some point I will also use the code in GSLIB to take
      advantage of some capabilities that are not included in Gstat.

      Hope this information was useful.


      Isobel Clark wrote,

      Dear Francisco

      The reason I developed the Geostokos software was
      because nothing else did what I needed to do.

      The reason we wrote Practical Geostatistics 2000 was
      because no-one had written a text book which could
      really be used for basic teaching.

      Check us out at http://geoecosse.bizland.com or



      Chandra S. wrote,

      Dear Colleague

      During 27 April to 04 May, I am on personal leave. I will respond to your
      message after I am back to office.

      Best regards

      S Chandra

      [Non-text portions of this message have been removed]
    Your message has been successfully submitted and would be delivered to recipients shortly.