Here is the JGR abstract style:

Abstract

The purpose of the abstract is twofold: (1) state the nature of the investigation and (2) summarize the important conclusions of this investigation. The abstract should be suitable for separate publication in an abstract journal and be adequate for indexing.

Set as a single paragraph
Limit to 250 words for JGR; 150 words for other journals.
Do not include references in the abstract.

No particular reason not to have multiple paragraphs; often I use multiple paragraphs when they say not to in AGU meeting abstracts, knowing that they don't correct them. But there are advantages of easier readibility if all papers use the same format.

As for the tenses, you probably know scientists are not great writers. Any one of several styles can be ok - the passive voice is appropriate in the sense that the facts are independent of the people writing, so why inject them? Or one can write like a history: I did this, I did that, I found this conclusion. Or one can write in relative terms: below, we will find some result, as we are finding here, and found in the last section.

The basic fact is that scientists care little for fancy grammar, good scientific writing is simple and transparent, and we use what works for us.

It's like the way we dress, we just don't care much. We are after how nature works, not polishing our verbs to a perfect sheen.