Ž . Global and Planetary Change 19 1998 87–114 Description and validation of the atmosphere–land–surface ž / interaction scheme ALSIS with HAPEX and Cabauw data Parviz Irannejad, Yaping Shao ) Centre for AdÕanced Numerical Computation in Engineering and Science, The UniÕersity of New South Wales, Sydney, NSW, Australia Received 12 September 1997; accepted 9 February 1998 Abstract Ž . A new land surface parameterization scheme ALSIS , with emphasis on soil moisture prediction, is described and validated with observations from HAPEX-MOBILHY and Cabauw. An important feature of the scheme is the inclusion of vertical heterogeneity of soil hydraulic parameters is modelling unsaturated flow. The simulated soil moisture for HAPEX site using a vertically homogeneous soil has a positive bias in the upper soil layers and a negative bias in the deep soil layers. Taking into account the soil vertical heterogeneity greatly eliminates this discrepancy and results in an excellent agreement between annual cycles of modelled and observed soil moisture profiles. The mean annual soil moisture in the top 1.6 m of soil increased from 394 mm for homogeneous case to 433 mm for the heterogeneous case, consistent with 435 mm observed. The improvement in soil moisture simulation resulted in an improved skill in predicting the mean and the diurnal Ž . cycles of surface fluxes for the intensive observational period 28 May–3 July . The simulated monthly averages of surface temperature and fluxes follow observations over the year, except for January when the model overestimates the latent heat flux due to its failure in simulating high rates of dew fall. The deviation of modelled monthly mean surface fluxes from observations are well within the estimated observational errors. The simulated mean daily surface temperature, and surface fluxes are generally consistent with observations, except for some times in the winter period. The modelled diurnal cycles of temperature and fluxes are in agreement with those observed. However, the model overestimates the night-time latent heat flux, especially during January. q 1998 Elsevier Science B.V. All rights reserved. Keywords: PILPS; land surface parameterization; soil moisture modelling; energyrwater balance; HAPEX-MOBILHY Cabauw; Cabauw 1. Introduction The land surface component of an atmospheric model should be able to realistically represent the partitioning of the surface available radiative energy between latent and sensible heat fluxes. Studies show that this Ž . partitioning has an important influence on the global climate e.g., Shukla and Mintz, 1982; Yeh et al., 1984 Ž . and on the mesoscale atmospheric circulation and precipitation Avissar and Pielke, 1989; Blyth et al., 1994 . The partitioning of the surface energy is closely linked to surface water balance through high dependency of surface evapotranspiration on the availability of soil moisture. Therefore, to provide a good approximation of ) Corresponding author. Fax: q61-2-9662-7792; e-mail: y.shao@unsw.edu.au 0921-8181r98r$ - see front matter q 1998 Elsevier Science B.V. All rights reserved. Ž . PII: S0921-8181 98 00043-5