Transfer projects

Over the last few years, considerable development of the ASCAT Wind Data Processor (AWDP), and in particular of the two-dimensional variational (2D-Var) ambiguity removal (AR) scheme, has been carried out. This has led to robust and consistent ASCAT Level 2 wind data products at various resolutions (25-km and 12.5 km) and sigma0 processing types, i.e., nominal (Hamming window averaging) and coastal (box-car averaging) processing. As smaller scales are revealed, the mesoscale analyses and consequently 2D-Var become more challenging. Nevertheless, a more detailed and case-study-oriented analysis has revealed that in certain conditions (see figure example), 2D-Var selects the wrong ambiguity along wind fronts, leading to noticeable misplacements of the front lines. Preliminary analysis shows that this is very likely induced by three factors: a) a wrong background field (i.e., when ECMWF fails to place the wind front in the correct location); b) ASCAT being a dual wind ambiguity system (in which, generally, two equally-likely solutions with similar wind speeds and opposed wind directions are obtained from inversion); c) an alignment (or close to alignment) of the erroneous NWP wind directions with the ASCAT wind ambiguities. When these three conditions occur, 2D-Var resolved wind front (convergence) is located where the wrong background (ECMWF) front is, since ASCAT wind ambiguities in themselves do not provide meaningful information to 2D-Var about where the convergence takes place (both ambiguities are equally likely). However, additional ASCAT information on the location of such wind disturbances is in principle available to help the 2D-Var process (see figure). This AS study proposal aims at the development of complementary information derived from the inversion and from an image processing technique (i.e., the so-called singularity analysis) to improve the current 2D-Var AR scheme for ASCAT in mesoscale conditions.