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

RE: AI-GEOSTATS: spatial sampling

Expand Messages
  • Rajive Ganguli, Ph.D., P.E., C.O.I
    Errors from a sufficiently large dataset will be normal even if the parent distribution is not normal (Stat. for Experimenters, Box et al). This is the
    Message 1 of 2 , Feb 27, 2004
    • 0 Attachment
      "Errors" from a sufficiently large dataset will be normal even if the parent
      distribution is not normal (Stat. for Experimenters, Box et al). This is
      the essence of CLT. However, as you hinted, if the autocorrelation is not
      take care of, then the two subsets will have different properties.




      Rajive Ganguli, Ph.D., P.E., C.O.I
      Associate Professor of Mining Engineering
      University of Alaska Fairbanks
      ================================
      Office: 317 Duckering Building
      Mailing Add: Box 755800, Fairbanks, AK 99775
      ph: 907-474-7631, fax: 907-474-6635
      web: http://www.faculty.uaf.edu/ffrg/
      -------------------------------------------
      "He uses statistics as a drunken man uses lamp-posts... for support rather
      than illumination." - Andrew Lang (1844-1912)




      > -----Original Message-----
      > From: ai-geostats-list@... [mailto:ai-geostats-list@...]On
      > Behalf Of Munroe, Darla K
      > Sent: Friday, February 27, 2004 8:27 AM
      > To: 'ai-geostats@...'
      > Subject: AI-GEOSTATS: spatial sampling
      >
      >
      > Hello,
      >
      > Please help me settle an argument with my co-author (and if I'm
      > wrong, I'll
      > graciously admit it!).
      >
      > We are talking about a regression analysis of land cover on a bunch of
      > geophysical and socioeconomic variables (topography, distance to roads,
      > market access, etc.). The objective is to get reliable estimates for beta
      > (or the average influence of each of the independent variables on the
      > dependent variable, land cover), irrespective of the spatial
      > relationships.
      >
      > To reduce spatial autocorrelation, we have been drawing regular spatial
      > samples, i.e., with a 5x5 pixel window, up to a 50x50 pixel window.
      >
      > First of all, due to the central limit theorem, if we have an adequately
      > large sample size given the data, two samples of differing window size
      > should yield the same betas and std errors, correct (assuming sampling has
      > effectively filtered spatial autocorrelation, which it may or may not)? I
      > think the central limit theorem is the justification for this assertion.
      >
      > Secondly, it should be empirically testable what the ideal
      > sampling distance
      > should be (using a semivariogram or the like) for this given data set?
      >
      > I would love to hear the list's perspectives on this issue.
      > - Darla Munroe
      >
      > --
      > * 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.