Evaluation of near-surface temperature, humidity, and equivalent temperature from regional climate models applied in type II downscaling

Atmosphere-surface interactions are important components of local and regional climates due to their key roles in dictating the surface energy balance and partitioning of energy transfer between sensible and latent heat. The degree to which regional climate models (RCMs) represent these processes with veracity is incompletely characterized, as is their ability to capture the drivers of, and magnitude of, equivalent temperature (T-e). This leads to uncertainty in the simulation of near-surface temperature and humidity regimes and the extreme heat events of relevance to human health, in both the contemporary and possible future climate states. Reanalysis-nested RCM simulations are evaluated to determine the degree to which they represent the probability distributions of temperature (T), dew point temperature (T-d), specific humidity (q) and T-e over the central U.S., the conditional probabilities of T-d|T, and the coupling of T, q, and T-e to soil moisture and meridional moisture advection within the boundary layer (adv(T-e)). Output from all RCMs exhibits discrepancies relative to observationally derived time series of near-surface T, q, T-d, and T-e, and use of a single layer for soil moisture by one of the RCMs does not appear to substantially degrade the simulations of near-surface T and q relative to RCMs that employ a four-layer soil model. Output from MM5I exhibits highest fidelity for the majority of skill metrics applied herein, and importantly most realistically simulates both the coupling of T and T-d, and the expected relationships of boundary layer adv(T-e) and soil moisture with near-surface T and q.