Direct detection Doppler wind lidar (DWL) has been demonstrated for its capability of atmospheric wind detection ranging from the troposphere to stratosphere with high temporal and spatial resolution. We design and describe a fiber-based optical receiver for direct detection DWL. Then the locking error of the relative laser frequency is analyzed and the dependent variables turn out to be the relative error of the calibrated constant and the slope of the transmission function. For high accuracy measurement of the calibrated constant for a fiber-based system, an integrating sphere is employed for its uniform scattering. What is more, the feature of temporally widening the pulse laser allows more samples be acquired for the analog-to-digital card of the same sampling rate. The result shows a relative error of 0.7% for a calibrated constant. For the latter, a new improved locking filter for a Fabry–Perot Interferometer was considered and designed with a larger slope. With these two strategies, the locking error for the relative laser frequency is calculated to be about 3 MHz, which is equivalent to a radial velocity of about 0.53 m/s and demonstrates the effective improvements of frequency locking for a robust DWL.
A mobile incoherent Doppler lidar was developed at the University of Science and Technology of China. The lidar consists of three subsystems. All subsystems are designed based on the well-proven double-edge technique, operate at 354.7 nm, and use Fabry–Perot etalons as frequency discriminators. The whole system is designed for wind measurement from 15- to 60-km height. In order to make the lidar receiver more compact and stable and to reduce interference between optical paths inside the receiver box, fiber splitters are introduced into the lidar receivers as a substitute for normally used discrete components. According to the stability of the splitter, the wind error dominated by the splitting ratio would be <0.49 m/s. To reduce luminance heterogeneity’s influence on the splitter performance, an integrating sphere is used in the system. Multiple measurements of transmission curves have a maximum mean squared error of 9.674E−5. A typical result of wind profile is also given to help demonstrate the reliability of the lidar and the fiber-based receiver.
A modification method is described for Rayleigh Doppler lidar wind retrieval. Compared to the double-edge theory of Korb et al. [Appl. Opt.38, 432 (1999)] and the retrieval algorithm of Chanin et al. [Geophys. Res. Lett.16, 1273 (1989)], it has a greater sensitivity. The signal-to-noise ratio of the energy monitor channel is involved in error estimation. When the splitting ratio of the two signal channels is 1.2, which usually happened during wind detection, it will improve the measurement accuracy by about 1% at 30 km altitude for a Doppler shift of 250 MHz (44 m/s ). Stabilities of retrieval methods, i.e., errors caused by the spectrum width deviation including laser pulse, Rayleigh backscatter, and filter transmission curve are first discussed. The proposed method increases the resultant precision by about 15% at 30-km altitude assuming an 8-MHz deviation in full width at half maximum of the Fabry–Perot interferometer.