One-dimensional soil moisture profile retrieval by assimilation of near-surface observations: a comparison of retrieval algorithms

This paper investigates the ability to retrieve the true soil moisture and temperature profiles by assimilating near-surface soil moisture and surface temperature data into a soil moisture and heat transfer model. The direct insertion and Kalman filter assimilation schemes have been used most freque ntly in assimilation studies, but no comparisons of these schemes have been made. This study investigates which of these approaches is able to retriev e the soil moisture and temperature profiles the fastest, over what depth s oil moisture observations are required, and the effect of update interval o n profile retrieval. These questions are addressed by a desktop study using synthetic data. The study shows that the Kalman filter assimilation scheme is superior to the direct insertion assimilation scheme, with retrieval of the soil moisture profile being achieved in 12 h as compared to 8 days or more, depending on observation depth, for hourly observations. It was also found that profile retrieval could not be realised for direct insertion of the surface node alone, and that observation depth does not have a signific ant effect on profile retrieval time for the Kalman filter. The observation interval was found to be unimportant for profile retrieval with the Kalman filter when the forcing data is accurate, whilst for direct insertion the continuous Dirichlet boundary condition was required for an increasingly lo nger period of time. It was also found that the Kalman filter assimilation scheme was less susceptible to unstable updates if volumetric soil moisture was modelled as the dependent state rather than matric head, because the v olumetric soil moisture state is more linear in the forecasting model. (C) 2001 Elsevier Science Ltd. All rights reserved.