There are two things driving the differences between these two texts: audience and format.
The article is a written text aimed at people with some expertise in atmospheric chemistry and the greenhouse effect. This means:
- It's assumed readers will have more background knowledge that they can use to understand special terminology, or unpack abstractions or summary.
- Expert readers often prefer a compressed style because it's easier to scan for terminology or key phrases to see whether the content is relevant to them.
- Readers can re-read anything that they don't understand straight away. (But you should still aim to be as clear as possible when writing!)
The speech is aimed at a novice audience who might not know much, if anything, about atmospheric stuff.
- Audience members are likely to have very little or no background knowledge. That means they will find it harder to unpack any terminology or abstractions.
- Because it's a live speech, listeners can't go back and re-listen to anything they didn't understand the first time. Shorter clauses and more connectors puts less load on the audience's working memory—they don't have to remember as many words at a time to make sense of what the speaker is saying. (Technically someone listening to the recording could go back and re-listen, but that's still a lot more fiddly than re-reading written work.)