There would seem to be a gap in the market for a set of broad climate change economic indices, i.e. based on prices or productivity, not meteorological data.

Go through the article by category: wheat, rice, water, residential property... Then pair locations in different climate contexts. For example: US property might use SF and Miami v. Boise and Nashville; wheat might use Nebraska v. Alberta. Combine the pairs to remove overall trends and leave differential performance (long-short).

Use existing tradeable indices for those local markets to construct global climate change index, perhaps rolled up in two different ways, by category and by country (or larger region). For example, global property, and everything in N.America.