You're now introducing circular references of arrays. Please get your understanding of a simple array straight first.

To add a column to an array, you use REDIM.

For example, an array with three rows and three columns is defined as DIM $a[2,2]. To add another columns to it, you'd do a REDIM $a[2,3], if you want to preserve the existing array element contents you'd use PRESERVE, thus REDIM PRESERVE $a[2,3]. Please experiment with this first to fully understand it. You now have a total of three rows and four columns. Arrays beyond two dimensions work the same way. Imagine an array as a collections ofindividual data elements, e.g. wooden cubes. A variable is a single wooden cube. A row or column of wooden cubes would represent a one-dimensional array, with each cube being a single data element. Now add a second ropw of wooden cubes to the first one, you now have a two-dimensional array of wooden cubes.

Then, put another two rows of cubes onto the existing ones, you now have a structure resembling a cube by itself, e.g. two cubes long, two cubes high, and two cubes in depth.

Now comes the tricky, or easy part. Instead of considering a single cube as a data element assume that the cube itself is made of a number of red wooden cubes, e.g. a 3x3x3 cube of red wooden cubes. This is an array-of-arrays.

I'm giving up now, I have no idea how to simplify it even more, maybe start with a drawing?

You MUST understand the basics of arrays before you can start using them effectively.
_________________________
There are two types of vessels, submarines and targets.