1
$\begingroup$

If we work on groundwater data, if the data is not normally distributed, can we apply geostatistics on this data without transforming it to normal distribution?

Improve this question
$\endgroup$
1
  • 1
    $\begingroup$ Hi Araminta. Welcome to CV. Is there any compelling reason why it has to be normal data and if other wise it has to be transformed? In general, normality can be a desirable trait, but lack of it is not a huge hindrance in itself. Please cite your sources that have prompted such assertion, if there is any. $\endgroup$
    – User1865345
    Commented Dec 1, 2023 at 7:01

1 Answer 1

Reset to default
2
$\begingroup$

Geostatistics is an entire branch of statistics -- it's all the statistics that are applied to spatial data. Some of its methods may require normally distributed data (see below), but many do not. For instance, kriging is a common geostatististical method and it does not require normal data, as this Wikipedia article makes clear.

Are you asking about some specific method within geostatistics?

Re Normality of data requirement: If you apply more general statistical methods to geographic data, I think that qualifies as geostatistics. But even, say, a linear multilevel model does not require normal data, it assumes normally distributed errors. And, if that assumption is grossly violated, you can use a nonlinear MLM.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.