Land surface heating and the North American monsoon anticyclone: Model evaluation from diurnal to seasonal

Data from several regional and global models (including model-based analysis data) are compared with field data from the North American Monsoon Experiment (NAME), from observational sites as well as satellite retrievals. On the regional scale (NAME tier 1.5), sensible heating is shown to exceed latent and is furthermore concentrated in the lower half of the troposphere, so in considering the North American monsoon (NAM) midlevel anticyclone, the authors focus on radiative and turbulent energy fluxes at the surface. Models exhibit large discrepancies in their simulation of the mean diurnal cycle of these fluxes as well as in their sensitivity of evaporative fraction to recent rainfall. Most of the models examined have too much net radiation due to excessive shortwave surface flux (too little cloud) and too much sensible heating. These high biases in sensible heating appear to drive overpredictions of both the daily and seasonal rise of 500-hPa heights in the NAM anticyclone. This diurnal-seasonal resemblance suggests that calibrating surface heating processes using readily field-observed diurnal variations could lead to improvements in seasonal-time-scale NAM simulations.