North America Hydroclimate Variability in CMIP5 Model Climate Simulations and Projections: Are Simulations Improving and Projections Converging?

Analysis of simulations and projections of regional hydroclimate variability over North America, especially the Central United States (U.S)., from the CMIP3 models indicate that the region imposes a notable challenge for global climate models. Difficulties arise due to the imperfect representation of processes that generate precipitation variability over the Central U.S. and the lack of consistent projections of tendencies and extremes in summer from the models analyzed. 

We propose a two tier analysis for the climate simulations and projections from the models participating in the new CMIP5 effort: An extensive evaluation and intercomparison of tendencies and extremes in projections from pentad to seasonal scales, and analysis of the mechanisms that generate summertime precipitation variability over the Central U.S.. Convergence of tendencies and extremes among the models will give some certainties to our societies, while a good assessment of the mechanisms in the models will give reassurance in the projections. The projections to be analyzed will be those emanated from the new scenario given by the Representative Concentration Pathway 8.5 which is characterized by increasing greenhouse gas emissions in the 21st century.

The analysis of mechanisms that generate precipitation variability over the Central U.S. in summer will be based of the analysis of the structure of the components of the atmospheric water balance in model simulations and projections. The main mechanism will be revealed by the relative magnitude of the regressions of regional precipitation indices on precipitation, moisture fluxes, and evapotranspiration; this analysis will be complemented with a correlation analysis of July precipitation with the previous months’ precipitation, to assess, indirectly, the strength of soil moisture feedback. Because observations indicate that precipitation variability is generated by remote forcing, two different versions of empirical function analysis (EOF) will be employed to detect the main modes of variability in the simulated and projected climates: 1) a rotated EOF analysis of summer SSTs and 700hPa heights, and an Extended EOF analysis of SSTs. Comparisons with observations, especially in near-term projections will reveal the realism of the mechanisms generating regional hydroclimate variability in the models.

The analysis of tendencies will include the whole region, however, the analysis of extremes and droughts will be focused on the Central U.S.,with secondary attention on Southwestern U.S.. Credibility in the projected tendencies will be given by their proximity with those observed at the end of the 20th century. The assessment of extremes will be done at pentad, monthly and seasonal resolutions, but the assessment of droughts will be done at low-frequency scales of seasonal data via histogram analysis.