I have a seismic error volume (standard deviation volume) in 3D. How can I generate an average error map (2D) between a time gate? I am thinking that averaging (like mean) of a standard deviation volume is not meaningful. Thank you in advance.
The following information is provided by Dr. P. Dewangan:
Yes, the combined means is easy to calculate than the combined
variance...There is a way to calculate the combined variance
(http://www.emathzone.com/tutorials/basic-statistics/combined-variance.html)
See the attached file for derivation of the formula.
Regards
Pawan
-----
Ramprasad
I'm not sure about the equation to solve the problem,...but if you're data really 3D data, so the distribution of your data on 3D with the distribution on 2D which is extracted from ur 3D will have the sama deviation,...but if your data was built from 2D data then you create 3D model, and you take 2D slice from them then it would be different story.
well,...that what I though,..
Thanks Ramprasad, Pawan and Siti. Let me rephrase my question. I have two 3D volumes of a rock property (say, seismic velocity or porosity) - 1) mean, 2) variance. So, for each point in 3D space I have an estimate (mean and variance) of a rock property. Many times we like to visualize them in a 2D (map), say an average between two time or depth levels. We can take a mean or other kind of averaging to produce a map from the mean estimate volume. We also like to know the error (variance) in a map sense. I want to know how people make an map from a variance volume. Thank you.
Owh,..I see,...as long as I know, to create a map from 3D volume with every point on the volume havd to e quantity value and error, so you need to give weight of your error to every direction,..for example. u have 3 coordinate x, y, z,..then u give ur error in every direction, like maybe on x direction u give error weight 1, error weight on y direction is 1, and error weight on z is 0.5,...when u extract map on direction z=a,...so the distribution of your value and their variance was function of (x,y,z=a).
may be that just little bit I know,...
- Badges
- Science topic
- Similar topics
- Mathematical Sciences
- Statistics
- Probability
- Probability Theory
- Reliability