Posters

Abstract

Dual Scanning Lidar (DSL) systems are increasingly utilized for characterizing offshore wind resources, offering the advantage of performing remote measurements offshore while being installed onshore, reducing logistical complexity, environmental impacts and increased safety. Their capacity to generate high-resolution wind field data over extended ranges makes them a compelling alternative to conventional measurements.  The most widely adopted scanning configuration is the point-intersect method, where laser beams from two time-synchronized scanning lidars converge at a fixed spatial coordinate to reconstruct virtual meteorological masts. Horizontal wind speed is derived from the individual line-of-sight (LOS) wind components aligned with each laser beam. The reliability of these LOS wind speed vectors directly affects the estimation of wind speed and turbulence parameters, which are critical for wind energy applications.     Obtaining reliable measurements from DSL systems depends critically on high combined data availability from both lidar units. Lidar measurements rely on the presence of sufficient aerosols in the air which serve as the backscatter target for the lidar beams.  Adverse atmospheric conditions such as low aerosol concentrations or overly clear air can significantly impair signal return, thereby reduce overall data quality or limit the measured range for DSL campaigns.  This study introduces a neural network–based reconstruction framework designed to mitigate data loss under low-aerosol conditions. The methodology is applied during a DSL campaign at Oldbaum’s Long-Range Lidar Test Facility (C-TEST) in Blyth, UK, where the lidars were deployed to reconstruct at an intersecting point 6.5km offshore from the lidar units, adjacent to an offshore met mast reference.  The neural-network method achieved substantial data recovery improvements over conventional techniques. When compared to the reference mast, the newly reconstructed wind fields demonstrated well-behaved uncertainty characteristics, consistent with theoretical lidar error models. These results validate the effectiveness of the proposed approach in enhancing DSL performance under challenging atmospheric conditions.