H31F-1490
Modelling evapotranspiration during precipitation deficits: identifying critical processes in a land surface model

Wednesday, 16 December 2015
Poster Hall (Moscone South)
Anna Ukkola1, Andy Pitman2, Mark R Decker3, Martin G De Kauwe4, Gab Abramowitz1, Yingping Wang5 and Jatin Kala6, (1)University of New South Wales, Sydney, Australia, (2)University of New South Wales, ARCCSS, Sydney, Australia, (3)University of New South Wales, Sydney, NSW, Australia, (4)Macquarie University, Sydney, Australia, (5)CSIRO, Ocean and Atmosphere Flagship, Aspendale, Australia, (6)Murdoch University, Murdoch, Australia
Abstract:
Surface fluxes from land surface models (LSM) have traditionally been evaluated against monthly, seasonal or annual mean states. Previous studies have noted the limited ability of LSMs to reproduce observed evaporative fluxes under water-stressed conditions but very few studies have systematically evaluated LSMs during rainfall deficits. We investigate the performance of the Community Atmosphere Biosphere Land Exchange (CABLE) LSM in simulating latent heat fluxes in offline mode. CABLE is evaluated against eddy covariance measurements of latent heat flux across 20 flux tower sites at sub-annual to inter-annual time scales, with a focus on model performance during seasonal-scale rainfall deficits. The importance of key model processes in capturing the latent heat flux is explored by employing alternative representations of hydrology, soil properties, leaf area index and stomatal conductance. We demonstrate the critical role of hydrological processes for capturing observed declines in latent heat. The effects of soil, LAI and stomatal conductance are shown to be highly site-specific. The default CABLE performs reasonably well at annual scales despite grossly underestimating latent heat during rainfall deficits, highlighting the importance for evaluating models explicitly under water-stressed conditions across multiple vegetation and climate regimes. A new version of CABLE, with a more physically consistent representation of hydrology, captures the variation in the latent heat flux during seasonal-scale rainfall deficits better than earlier versions but remaining deficiencies point to future research needs.