Impact of vertical and temporal error covariances on the vertical velocity and advective tendencies in the ARM constrained variational analysis

Description

Deriving the atmospheric vertical velocity and advective tendencies accurately from the ARM field campaigns is essential to force and evaluate models against measurements, especially measurements of clouds and precipitation. In the current ARM analysis, a variational algorithm is used to derive these physical quantities. The magnitudes of the variational adjustments depend on the specified error covariance matrix in the definition of the minimization cost function. If the specified error is large for a variable at a particular level, the adjustment to that variable at that location will be large. At present, errors for all atmospheric state variables are specified using instrument errors from manufacturers and aliasing errors from temporal variation of the variables, but they are assumed to be independent in time and in space. In this presentation, we show the structure of the error covariance matrix in the TWP-ICE field campaign and quantify its impact on the analyzed products. Simulations from a high-resolution model (WRF) are used to verify the implementation of the algorithms and results.