Sub-Seasonal Predictability of North American Monsoon Precipitation

North American Monsoon (NAM) rainfall is a vital water resource in the United States Southwest, providing 60-80% of the region's annual precipitation. However, NAM rainfall is highly variable and water managers lack skillful guidance on summer rainfall that could help inform their management decisions and operations. Here we show that NAM season (June-October) precipitation can be forecasted by the European Centre for Medium-Range Weather Forecast's model months ahead at catchment scales. This is possible by identifying the frequency of days with synoptic-scale moisture advection into the NAM region, which greatly improves predictability over directly utilizing modeled precipitation. Other forecasting systems fail to provide useful guidance due to deficiencies in their data assimilation systems and biases in representing key synoptic features of the NAM including its teleconnections.

Automated summary

North American Monsoon (NAM) rainfall, which contributes up to 80% of the annual precipitation in the United States Southwest, is highly variable and lacks skillful predictions. However, the European Centre for Medium-Range Weather Forecast's model can forecast NAM season precipitation months in advance by identifying the frequency of days with synoptic-scale moisture advection into the NAM region, improving predictability compared to other systems.