Best approach to merging two tables together (one really big, one really small)

We have a situation where we have a large fact table (approximately 1 billion rows, partitioned) that represents a sort of "baseline" set of data. We have another fact table with the same structure that collects "adjustments" to this same data. For business reasons, we need to keep them separate.

The "adjustment" table is very small, roughly 50k rows and each row in this table essentially replaces exactly one of the rows in the larger table. Also, in some cases we can have rows in the adjustment table for which there is no corresponding rows in the baseline. Obviously, the reverse of that is also true, most baseline rows are NOT adjusted.