The Role of Precipitation Recycling in the Propagation and Intensification of Droughts in North America

Predicting droughts allows stakeholders to mitigate some of the negative impacts of these natural disasters. However, there are still large gaps of knowledge regarding the physical drivers of drought onset, development, and recovery. These gaps have limited our ability to predict some important droughts and to understand how they may be affected by climate change. One physical mechanism that has been linked to the evolution of droughts is precipitation recycling, but its role has not been quantified in detail. Here we use a precipitation recycling model that backtracks the spatial origins of precipitation using vertically integrated moisture fluxes and evapotranspiration data. This allows us to estimate the climatology of moisture sources and sinks, and identify from where moisture fails to arrive when a given region experiences a drought. ERA-Interim data is used to drive this precipitation recycling model from 1980 to 2016 throughout North America and its surrounding oceans. The climatological analysis shows that oceans contribute around 80% of the precipitation over North America during winter, while precipitation that originates from evapotranspiration over land reaches a relative contribution of 60% in the summer. Precipitation contributions from the Pacific Ocean were found to be significantly and positively correlated with ENSO and PDO indices. Furthermore, a regression analysis showed that dry soil moisture in the US Southwest reduces moisture exports to the US Midwest, which in turn can dry soil moisture in the US Midwest. Given that up to 13% of precipitation over the US Midwest was found to be locally recycled, there is a multiplier effect whereby a 10 mm/month reduction in precipitation imports into the region leads to an additional decrease of 0.8 mm/month (on average) from reduced local precipitation recycling, causing a drought to intensify. It was also found that during extensive droughts (e.g. 2011 in Texas and 2012 in the US Midwest), precipitation deficits from land were as large and sometimes larger than those from the oceans. These results demonstrate a mechanism by which droughts can propagate spatially downwind, and point out to processes that need to be captured accurately by models used in operational seasonal forecasting and in climate change studies.