Publications & Events
Ensemble transform sensitivity method for adaptive observations
Abstract
The Ensemble Transform (ET) method has been shown to be a useful approach to provide guidance for adaptive observation deployment. It predicts forecast error variance reduction for each of possible deployments using its corresponding transformation matrix in an ensemble subspace. In this paper, an ET-based sensitivity (ETS) method, which calculates the gradient of forecast error variance reduction to analysis error variance reduction, is proposed to specify regions for possible adaptive observations. ETS is a first order approximation of the ET, but only needs one single calculation of transformation matrix, and increases computation efficiency. It is also found that using the original transformation matrix formulation, ETS is not sensitive to large analysis error variance in ensemble subspace. A new transformation matrix formulation is introduced, and the sensitivity of a new ETS based on this transformation is explicitly proportional to the analysis error variance. Using the new transformation matrix formulation, areas with larger analysis error variances are more likely identified as sensitive regions, which reflects analysis error variance impact in more meteorologically and statistically meaningful manners. In this paper, the two advantages of a newly formed ETS are presented and both of ET and ETS methods are applied to a real hurricane Irene (2011) case for comparison. The results confirm the improvements.
