Robust Changes in North America's Hydroclimate Variability and Predictability

Climate change adaptation planning requires a robust understanding of the projected change in hydroclimate variability and predictability. We use two large ensemble data sets to quantify the projected change in land hydroclimate variability and its potential predictability. Additionally, we use a reddened El Nino-Southern Oscillation (ENSO) framework that partitions annually averaged root-zone soil moisture variability into ENSO, land surface memory, and internal variability to understand drivers of the changes in hydroclimate variability and predictability. Even when global warming is projected to increase ENSO and its teleconnected precipitation variability over North America, we find that the corresponding change in soil moisture variability is relatively small and even decreases. This counter-intuitive result occurs since there is also a concurrent projected reduction in land surface memory due to global warming, which leads to reduced year-to-year persistence of soil moisture variability. Further, we find that regional mean state land surface (soil moisture) changes primarily drive future drought and pluvial risks, suggesting that infrastructure planning can incorporate robust mean state changes despite uncertainty in the variability projections. For the regions and in the models where the ENSO signal increases, we also find a concomitant shift in the frequency of drought and pluvial events, with higher power on inter-annual time scales but less power on decadal time scales, enhancing inter-annual hydroclimate predictability.

Automated summary

Adaptation planning for climate change requires a comprehensive understanding of hydroclimate variability and predictability. Using large ensemble data sets, we quantify the projected change in land hydroclimate variability and its predictability. We find that while global warming intensifies El Nino-Southern Oscillation (ENSO) and its associated precipitation variability, the corresponding change in soil moisture variability is relatively small or even decreases due to the concurrent reduction in land surface memory. Regional mean state changes in land surface (soil moisture) primarily drive future drought and pluvial risks, suggesting that infrastructure planning can incorporate robust mean state changes despite uncertainties in variability projections. In regions where the ENSO signal increases, we observe a shift in the frequency of drought and pluvial events, with higher power on inter-annual time scales but less power on decadal time scales, enhancing inter-annual hydroclimate predictability.