On wikipedia, I see lognormal distribution is defined as:
but in a paper I feel that it is different from the above wiki definition. Please see the below attached image. The exponential '2' is marked with a red circle. The extra '2' in equation (4) makes it different from the wiki definition. Thus, it is not a log normal distribution (for the blue underline marked) any more.
If the article is correct, then the square of a Gaussian variable is still a Gaussian. Do you think so? I do not hear about this.
What can be deduced from the article is not that the square of a Gaussian variable is still a Gaussian, but that the square of a log-normal variable is still log-normal.