Assessing GFDL high-resolution climate model water and energy budgets from AMIP simulations over Africa

Abstract

This study assessed surface water and energy budgets in Atmospheric Model Intercomparison Project (AMIP) simulations of a coupled atmosphere‐land model developed by Geophysical Fluid Dynamics Laboratory (Atmospheric General Circulation Model (AM2.5)). The AM2.5 water and energy budget variables were compared with four reanalyses data sets and an observational‐based reference, the Variable Infiltration Capacity model simulations forced by Princeton Global Meteorological Forcing (PGF/VIC) over 20 year period during 1991–2010 in nine African river basins. Results showed that AM2.5 has closed water and energy budgets. However, the discrepancies between AM2.5 and other data sets were notable in terms of their long‐term averages. For the water budget, the AM2.5 mostly overestimated precipitation, evapotranspiration, and runoff compared to PGF/VIC and reanalyses. The AM2.5, reanalyses, and PGF/VIC showed similar seasonal cycles but discrepant amplitudes. For the energy budget, while the AM2.5 has relatively consistent net radiation with other data sets, it generally showed higher latent heat, lower sensible heat, and lower Bowen ratio than reanalyses and PGF/VIC. In addition, the AM2.5 water and energy budgets terms mostly had the smallest interannual variability compared to both reanalyses and PGF/VIC. The spatial differences of long‐term mean precipitation, runoff, evapotranspiration, and latent heat between AM2.5 and other data sets were reasonably small in dry regions. On average, AM2.5 is closer to PGF/VIC than R2 and 20CR are to PGF/VIC but is not as close as Modern‐Era Retrospective analysis for Research and Applications and Climate Forecast System Reanalysis to PGF/VIC. The bias in AM2.5 water and energy budget terms may be associated with the excessive wet surface and parameterization of moisture advection from ocean to land.