On a new method of determining the height of the nightglow