Retrieval of Moisture from GPS Slant-path Water Vapor Observations using 3DVAR with Isotropic and Anisotropic Recursive Filters

It is well known that flow-dependent background error covariance plays an important role in data assimilation. Such error covariance is usually spatially inhomogeneous and anisotropic. In our recent study (Liu and Xue 2005), a three-dimensional variational (3DVAR) analysis system is developed that models the flow-dependent background error field using an explicit spatial filter. Better moisture analysis from simulated GPS slant-path water vapor data is obtained when using an anisotropic spatial filter based on the flow-dependent background error. The explicit filter, when applied over even a moderate number of grid points in three dimensions, is still expensive in terms of both computational and memory storage costs. A much more computationally efficient algorithm is the implicit recursive filter, which is, on the other hand, complex to implement, especially for the anisotropic one.

In this paper, we implement the recursive filter with an anisotropic option in our 3DVAR analysis system. We perform retrieval experiments of GPS slant-path water vapor data and compare the results with those of explicit filters. It is found that the analyses with recursive filters are at least as good as those with corresponding explicit filters and in some cases the recursive filter performs even better. Improved positive-definiteness property of the modeled background error covariance is believed to be the reason.