SRTM Forum
Welcome to the SRTM Forum.  This is a discussion forum for users of data from the Shuttle Radar Topography Mission.
 


Return to Website

  Reply
  Home

Subject:   Re: absolut error in accuracy report
Name:   Vladimir Stoje
Date Posted:   Jan 26, 07 - 12:27 PM
Email:   stoje@volny.cz
Message:   Hi Carlos,

If I unerstand this report correctly (with my bad English, sorry), the described errors are always 90% errors, so the interval plus minus E covers 90% of values. The interval plus minus SDT covers 68% of values and in the normal distribution the SDT should be 1.64 smaller as the 90% Error, for example if the absolute height error for Eurasia is about 6.2m, in the case of normal distribution the SDT should be 6.2/1.64 = cca 3.8m. There are many diagrams and tables described the actual 90% errors and SDT, and the ratio isnt exactly 1.64, because the observed distributions are not precisely normal. I hope this is 90% correct...

Cheers

Vladimir
Replies:    
Re: Re: absolut error in accuracy report by Vladimir Stoje · Jan 26, 07 - 12:46 PM
Re: Re: absolut error in accuracy report by Carlos Grohmann · Jan 28, 07 - 1:07 PM


  Reply
  Home


powered by Powered by Bravenet bravenet.com