generative nowcasting
One of 50 themes found in 20,416 EGU water abstracts (2024 to 2026). Its most distinctive words are below — they are what separates this body of work from the rest of the corpus.
How to read this theme
This theme holds 356 of the 20,416 water abstracts, making it the 33th-largest of the 50 themes. Between 2024 and 2026 it gained 0.88 points of share.
96.1% of its abstracts were flagged as using machine learning, 82.9 points above the 13.2% corpus average. AI is already the dominant method here, so novelty is more likely to come from the science than from the model.
This theme is one of the 4 whose rise survives the trend test described in the methodology.
Find data for this theme
Search the directory for datasets matching this theme's defining terms.
Semantically nearest themes
Themes that sit closest to this one on the semantic map — the work that reads most like it.
Themes are clusters found by grouping abstracts on their sentence embeddings, so their boundaries are statistical rather than official EGU categories. "Uses AI" comes from a keyword detector validated at precision 0.95 and recall 0.9. Point estimates carry roughly half a point to a point of uncertainty from the clustering. Full method on the main trends page.