MASDA - MODAS Adaptive Sampling Decision Aid

Abstract

The Modular Ocean Data Assimilation System (MODAS) produces oceanographic nowcasts based on (a) climatology, (b) remotely-sensed sea surface temperature and height, and (c) in situ measurements. Recent analyses have shown that the locations of in situ measurements can have a profound influence on the accuracy of the MODAS synthetic profiles. Small-scale variability combined with sparse sampling and inappropriate covariance scales can lead to a spreading of unrepresentative anomalies. The MODAS Adaptive Sampling Decision Aid (MASDA) is being developed to guide the selection of airborne expendable bathythermograph (AXBT) measurement locations to improve the accuracy of MODAS analyses while minimizing the number of required measurements. MASDA uses the computed MODAS temperature uncertainty to predict the optimum sampling locations. The iterative in-flight MASDA approach is to recommend sequential measurement locations based on sequentially computed temperature uncertainty. The pre-flight combinatorial MASDA approach is to recommend the best combination of N measurement locations based on the computed temperature uncertainty. These environmentally driven sampling strategies are expected to increase accuracy of MODAS analyses relative to MODAS analyses based on alternate sampling strategies with the same number of observations. AXBT measurements from several ocean areas are being used to develop and test MASDA algorithms. Preliminary results showing improvement in MODAS accuracy using the MASDA method for selecting observations compared to more subjective selection methods are presented.