Drought changes and the mechanism analysis for the North American Prairie

Abstract

The worst droughts in the central part of the North American Prairie in the past several hundred years have been reconstructed from tree-ring chronologies, suggesting that some drought years have exceeded the severity shown by the gauge record. A general circulation model of the Geophysical Fluid Dynamics Laboratory (GFDL) has simulated climate changes for the area during the past 250 years driven by climatic forces, providing scenarios of extreme climate that can further diagnose the mechanisms. This study refined the drought signals from the tree ring data and GFDL modeling at inter-annual and decadal time scales and analyzed the potential mechanisms driving the droughts. Results showed that drought years with summer precipitation lower than the 10th percentiles occurred during 1777–1789, 1847–1861 and 1886–1879 AD in the area. Both tree rings and model revealed that the frequency of droughts has been relatively consistent in a similar timing and frequency with climate change. Monte Carlo analysis have detected that the tree ring chronologies have recorded drought years with probabilities of 9.3%–12.8%, and the model has simulated the droughts with probabilities 5.7%–17.8%. Under CO2 and aerosol forcing, the GFDL modeled the drought recurrences of 13 years and 25 years, which are very synchronous changes with tree rings and consistent with gauge records. The 20-a and 10-a time scale reoccurrences of droughts are very consistent with solar radiation cycles, and similar to the length of cycles in oceanic records, suggesting that terrestrial precipitation modeling is properly driven from sun-land-sea dynamics. Detected severity, variability and return periods of droughts from the present study make potential improvements in drought predictions and constructing scenarios for climate impacts and adaptation strategies.