Comparison of latent heat flux estimates over Australia

Documenting the water cycle through modelling and observation is needed for more fundamental understanding of water exchanges in Earth's coupled systems. Spatially and temporally consistent characterization of the hydrological cycle has been a research goal of numerous investigations for many years. Recent applications of sophisticated modelling approaches has seen the development of a number of continental and global scale land surface schemes designed for this purpose, providing insight into the movement of water through the terrestrial system. The Global Land Data Assimilation System (GLDAS) is one such programme that has delivered significant insights into water and energy exchange over the Earth's terrestrial surfaces. While there are a number of numerical modelling based approaches that seek to describe terrestrial water and energy cycles, an operational, observationally based and temporally consistent data set for continental scale evapotranspiration is not currently available. As a result, important insights available from such data do not contribute to model assessment and calibration. This study compares latent heat fluxes derived from two land surface models, which have been run over Australia for a 2 year period, with satellite observations and in-situ rainfall measurements. The two land surface models, component schemes of the Global Land Data Assimilation System, are evaluated against a newly developed remote sensing based flux dataset, to identify their degree of hydrological coherence. The remote sensing product utilizes multiple satellite sensors onboard NASA's Aqua satellite to estimate instantaneous heat fluxes at the time of satellite overpass, providing an unprecedented spatial coverage of continental scale evaporation over Australia. An assessment of the agreement between these distinct data sets is undertaken, with a focus on the reproduction of spatial and temporal patterns of latent heat flux over portions of the Australian continent.