Comparing Evaporative Sources of Terrestrial Precipitation and Their Extremes in MERRA Using Relative Entropy

摘要

A quasi-isentropic, back-trajectory scheme is applied to output from theModern-Era Retrospective Analysis forResearch andApplications (MERRA) and a land-only replaywith corrected precipitation to estimate surface evaporative sources of moisture supplying precipitation over every ice-free land location for the period 1979- 2005. The evaporative source patterns for any location and time period are effectively two-dimensional probability distributions. As such, the evaporative sources for extreme situations like droughts or wet intervals can be compared to the corresponding climatological distributions using the method of relative entropy. Significant differences are found to be common and widespread for droughts, but not wet periods, when monthly data are examined. At pentad temporal resolution, which is more able to isolate floods and situations of atmospheric rivers, values of relative entropy over NorthAmerica are typically 50%-400% larger than atmonthly time scales. Significant differences suggest that moisture transport may be a key factor in precipitation extremes. Where evaporative sources do not change significantly, it implies other local causes may underlie the extreme events.