On the following page how is log normal distribution used. Also the pdf defined looks like log(log normal pdf) though not complete since if you see the wikipedia page then there's one x at the bottom too. It's confusing. Can someone refer the paper they are using? Why is the model being used to compute logvariance. Any particular reasons? No article on the internet talks about lognormal distribution and how the math is being done.