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

AI-GEOSTATS: bad regression between predicted and measured

Expand Messages
  • rocchini@unisi.it
    I ve made an interpolation by IDW, Spline and ordinary kriging and a cross-validation for all methods. When I make a regression between predicted and measured
    Message 1 of 3 , Oct 13, 2003
    • 0 Attachment
      I've made an interpolation by IDW, Spline and ordinary kriging and
      a cross-validation for all methods.
      When I make a regression between predicted and measured values I reach an R2
      of 0.0109 (spline), 0.006 (kriging), 0.0006 (idw), namely very flat curves
      (lines).
      What could I do? I think I'm using a small number of input values (62) for
      the area that I'm considering (4000 km2).
      Thanks
      Duccio Rocchini

      Dr. Duccio Rocchini
      Dpt.of Env. Sci.
      University of Siena
      www.rocchini.net





      --
      * To post a message to the list, send it to ai-geostats@...
      * As a general service to the users, please remember to post a summary of any useful responses to your questions.
      * To unsubscribe, send an email to majordomo@... with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list
      * Support to the list is provided at http://www.ai-geostats.org
    • Isobel Clark
      Duccio There are many reasons why your interpolations may not be working. A few of these are: # you are beyond the range of influence of any distance
      Message 2 of 3 , Oct 13, 2003
      • 0 Attachment
        Duccio

        There are many reasons why your interpolations may not
        be working. A few of these are:

        # you are beyond the range of influence of any
        distance relationship, that is you have too widely
        spaced sampling.

        # your data may have a skewed or other non-Gaussian
        distribution which makes both semi-variogram and
        correlation calculations invalid

        # you may have discontinuities, trends or anisotropies
        which have not been factored into your model

        Isobel
        http://drisobelclark.ontheweb.com

        ________________________________________________________________________
        Want to chat instantly with your online friends? Get the FREE Yahoo!
        Messenger http://mail.messenger.yahoo.co.uk

        --
        * To post a message to the list, send it to ai-geostats@...
        * As a general service to the users, please remember to post a summary of any useful responses to your questions.
        * To unsubscribe, send an email to majordomo@... with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list
        * Support to the list is provided at http://www.ai-geostats.org
      • Russell Barbour
        Dear Dr. Rocchini: Well, my first reaction is that R squared is not an appropriate measure for correlated data. Secondly, such a large area may well have an
        Message 3 of 3 , Oct 14, 2003
        • 0 Attachment
          Dear Dr. Rocchini:

          Well, my first reaction is that R squared is not an appropriate measure for
          correlated data. Secondly, such a large area may well have an underlying trend
          which should be removed before kriging is attempted and lastly the number of
          observations seems very small if that is the only data layer available.

          Sincerely

          Russell Barbour Ph.D.
          Research Associate in Applied Mathematics
          Vector Ecology Laboratory
          Yale School of Medicine
          60 College St. Rm 600
          New Haven CT. 06520
          TEL: 203 785 2394
          FAX 203 785 3604
          email: russell.barbour@...


          Quoting rocchini@...:

          > I've made an interpolation by IDW, Spline and ordinary kriging and
          > a cross-validation for all methods.
          > When I make a regression between predicted and measured values I reach an R2
          > of 0.0109 (spline), 0.006 (kriging), 0.0006 (idw), namely very flat curves
          > (lines).
          > What could I do? I think I'm using a small number of input values (62) for
          > the area that I'm considering (4000 km2).
          > Thanks
          > Duccio Rocchini
          >
          > Dr. Duccio Rocchini
          > Dpt.of Env. Sci.
          > University of Siena
          > www.rocchini.net
          >
          >
          >
          >
          >
          > --
          > * To post a message to the list, send it to ai-geostats@...
          > * As a general service to the users, please remember to post a summary of any
          > useful responses to your questions.
          > * To unsubscribe, send an email to majordomo@... with no subject and
          > "unsubscribe ai-geostats" followed by "end" on the next line in the message
          > body. DO NOT SEND Subscribe/Unsubscribe requests to the list
          > * Support to the list is provided at http://www.ai-geostats.org
          >

          --
          * To post a message to the list, send it to ai-geostats@...
          * As a general service to the users, please remember to post a summary of any useful responses to your questions.
          * To unsubscribe, send an email to majordomo@... with no subject and "unsubscribe ai-geostats" followed by "end" on the next line in the message body. DO NOT SEND Subscribe/Unsubscribe requests to the list
          * Support to the list is provided at http://www.ai-geostats.org
        Your message has been successfully submitted and would be delivered to recipients shortly.