Evaluation of Aeolus wind observations and their contribution to surface wind forecasts

Wind profile information is critical for understanding atmospheric dynamics and improving numerical weather prediction (NWP), which in turn benefits our daily activities and supports decision-making in a variety of industries, such as renewable energy, aviation, and ocean shipping. However, the absence of distributed wind profile measurements, particularly in regions over the oceans, tropics, and Southern Hemisphere, is one of the major deficiencies of the Global Observing System. To fill this gap, the European Space Agency launched the Aeolus satellite in August 2018. Owing to the state-of-the-art onboard Doppler wind lidar (DWL), the vertical profile of the horizontal line-of-sight (HLOS) wind velocity can be determined based on the Doppler shift of light backscattered by air molecules and particulates moving with winds. Before wind profile data can be utilized for practical applications, such as in the wind energy sector, comprehensive assessments of their error characteristics and their value to NWP model performance are required.

In light of the above, this PhD study aims to evaluate the performance of the space-borne DWL in wind profile detection and its contribution to near-surface wind forecasts. This study is scoped by three objectives. The first objective is to validate the wind profiles of Aeolus over Australia, which complements the validation in the Southern Hemisphere. The second objective is to investigate the impact of Aeolus wind profile assimilation on global sea surface wind forecasts based on the observing system experiments conducted by the European Centre for Medium-Range Weather Forecasts (ECMWF) comparing to surface wind observations from satellite scatterometers and ocean buoys. The third objective is to assess Aeolus' impact on near-surface wind forecasts over high-latitude lands in both hemispheres comparing to weather station data. Inter-comparison and triple collocation analyses were performed to achieve these objectives.

By taking wind profile measurements from ground-based wind profiling radars (WPRs) as the reference data, the Aeolus wind profile validation in Australia demonstrates that both Rayleigh- clear and Mie-cloudy winds achieve the mission bias requirement of 0.7 m s^-1. Mie-cloudy winds are more precise than Rayleigh-clear winds, with random errors of 4.14 m s^-1 and 5.81 m s^-1, respectively. Similar precisions are obtained from the triple collocation analysis based on the Aeolus, WPR, and ECMWF model data. In addition, the inter-comparison analysis shows that Mie-cloudy winds have smaller bias and higher precision than Rayleigh-clear winds at altitudes below 1,500 m, suggesting a greater influence of Mie-cloudy than Rayleigh-clear winds on data assimilation for NWPs in the planetary boundary layer.

Verifications based on the ECMWF Aeolus impact experiments and ocean surface wind observations indicate that Aeolus benefits the short-range (within 12 hours) sea surface wind forecasts for the global ocean, except in the tropical regions. Moreover, Aeolus can reduce the large-scale zonal biases of the short-range forecasts, whereas its impact on meridional biases varies with regions. As the forecast step is extended to T+120 h, the positive impact of Aeolus becomes more evident, particularly for the extratropical ocean regions in the Southern Hemisphere. For high-latitude lands, the positive contribution of Aeolus is mainly observed in the Northern Hemisphere after T+120 h, during the period with good Aeolus data quality, and during boreal winter and stormy conditions.

The research conducted in this PhD study extends our understanding of Aeolus observations and their performance in wind profile detection and their contribution to surface wind forecasts over the global ocean and over land at high latitudes in the Northern and Southern Hemispheres.