The North American Monsoon (NAM) is essential for the hydrology of the US Southwest and large parts of Mexico. In New Mexico, for example, the NAM contributes up to 60 % of the state’s total annual precipitation. With progressing climate change, monsoon precipitation might become even more important since spring and summer runoff from melting snowpack is decreasing with warming. This study identifies four hydrologically important weather types (WTs) during the North American warm season (June to October) and aims to understand how changes in these WTs contributed to changes in precipitation. Two of these WTs resemble NAM conditions with one bringing moderate precipitation from the Atlantic and Pacific whereas the other advects predominantly Gulf of Mexico moisture and leads to high precipitation totals. The other two WTs are predominantly dry.

In New Mexico, observations show a ~5 % per decade drying from 1980 to 2010. This is caused by a combination of increasing dry WT frequencies and a decrease in precipitation rates. Using 40 members of a perturbed initial condition global climate model ensemble shows that the main cause of observed WT frequency changes is natural climate variability and that anthropogenic climate change played a minor role. Analyzing simulated WT frequency changes from 1920 to 2100 (2006 to 2100 assumes a business as usual, RCP8.5, climate change forcing) shows a systematic frequency decrease in the monsoonal flow from 1920 to 2045 and a rapid increase thereafter. This results in a significant increase in warm season precipitation frequency by the end of the century, which is, however, counteracted by decreases in precipitation rates. The large-scale drivers, implications, and uncertainties of these results will be discussed.