Assimilation of semi-qualitative observations with a stochastic Ensemble Kalman Filter

TL;DR

This paper introduces the EnKF-SQ, a novel data assimilation method that effectively incorporates semi-qualitative observations with detection limits into ensemble Kalman filtering, improving forecast accuracy in linear and non-linear models.

Contribution

The paper develops the EnKF-SQ, a new scheme for assimilating out-of-range observations with detection limits, enhancing data assimilation capabilities over existing methods.

Findings

EnKF-SQ outperforms traditional methods in twin experiments.

Assimilating qualitative data improves forecast mean accuracy.

Performance depends on ensemble size and observation detection limits.

Abstract

The Ensemble Kalman filter assumes the observations to be Gaussian random variables with a pre-specified mean and variance. In practice, observations may also have detection limits, for instance when a gauge has a minimum or maximum value. In such cases most data assimilation schemes discard out-of-range values, treating them as "not a number", at a loss of possibly useful qualitative information. The current work focuses on the development of a data assimilation scheme that tackles observations with a detection limit. We present the Ensemble Kalman Filter Semi-Qualitative (EnKF-SQ) and test its performance against the Partial Deterministic Ensemble Kalman Filter (PDEnKF) of Borup et al. (2015). Both are designed to explicitly assimilate the out-of-range observations: the out-of-range values are qualitative by nature (inequalities), but one can postulate a probability distribution for them and then update the ensemble members accordingly. The EnKF-SQ is tested within the framework of twin experiments, using both linear and non-linear toy models. Different sensitivity experiments are conducted to assess the influence of the ensemble size, observation detection limit and a number of observations on the performance of the filter. Our numerical results show that assimilating qualitative observations using the proposed scheme improves the overall forecast mean, making it viable for testing on more realistic applications such as sea-ice models.