BLUE for the Scatterometer Constellation
High-resolution satellite-derived sea-surface wind data, such as those from scatterometers, are increasingly available for operational monitoring and forecasting of the ocean. In this presentation we present the scatterometer constellation and show first results that indicate that the scatterometer wind data assimilation process can be potentially much improved by bias correction.
Belmonte & Stoffelen (2019) documented large NWP model errors on both large and small scales and on all temporal scales. Trindade et al. (2019) succeeded to compute stable gridded NWP minus scatterometer differences over a few days, which were subsequently used to correct a time series of numerical weather prediction (NWP) outputs in a product that they called ERA*. They verified ERA* against independent wind measurements, which resulted in a reduction of the observation minus NWP field, so-called o-b, global variance of 20%. An o-b reduction of 20% is very substantial. Moreover, the reduction is due to bias correction, which helps to more closely fulfill the BLUE paradigm of NWP data assimilation. BLUE stands for Best Linear Unbiased Estimate and is necessary to well inform the atmospheric state variables of the NWP model.
Data assimilation of scatterometer data in NWP generally only resolves NWP spatial scales of ~150 km. Therefore, information on the wind-SST interaction, the diurnal wind cycle and the wind variability in moist convection areas is lost in such products. Moreover, known systematic NWP model (parameterization) errors and ocean currents contribute to local biases, which are rather time invariant. Such biases prevent to properly correct the meso- and large-scale NWP dynamical errors. Following ERA*, a bias correction scheme may be implemented that takes out the slow biases in o-b mentioned above, that are mostly related to the local ocean state and not resolved in NWP models.
The corrections can be further used to attribute errors to the bias (and variability) source terms in order to improve the air-sea interaction modelling, which will be important in the coupling of atmosphere and ocean models of the future.
