Tibetan is usually called an alphabet, and the atlas calls it one, but the technical description is an abugida, and the difference is worth stating because it explains several things that otherwise look arbitrary.
In an alphabet, consonants and vowels are letters of the same kind and are written in sequence. In an abugida, a consonant sign carries an inherent vowel, and a different vowel is written by modifying the consonant rather than by adding a letter beside it. Tibetan works the second way. The bare sign ka is read ka, not k, and the four vowel signs replace the inherent a rather than supplying a vowel where there was none. Three of them are written above the consonant and one beneath it, and none of them can stand alone.
The consequences run through the whole script. A syllable is a two-dimensional assembly rather than a line of letters, which is why it stacks and why block printing suited it. The unit of writing is the syllable rather than the word, which is why the tsheg separates syllables and why Tibetan has no word space. And a vowel-initial syllable needs a carrier, which is why a chen exists at all.
The Indic scripts of the Brahmi family are all built this way, and Tibetan inherited the design along with the letters.