This PhD thesis builds around a light source forming the basis for a novel type of wind measuring lidar. The lidar emits a train of laser pulses with each pulse being separated from its neighbours in frequency, while being closely spaced in time, thus combining the advantages of conventional continuous wave (CW) and pulsed lidars. A light source capable of emitting such a pulse train is suggested. A theoretical description of all components constituting the light source is presented, and a time dependent model is developed and compared to measurements as well as to previous theoretical work from the scientific literature. The model presented shows good agreement with the experimental results regarding the pulse train envelope as well as the individual pulses. A model adopted from the literature is subsequently expanded to incorporate frequency components other than the main signal frequency and compared to measurements of individual pulse spectra. Critical issues such as various contributions to noise, in particular amplified spontaneous emission (ASE), are investigated. The realized frequency stepped pulse train (FSPT) emitting light source has been incorporated into a modified CW lidar, and the ability to measure wind speeds as well as the direction successfully demonstrated. A challenge still remains in the improvement of the signal to noise ratio (SNR), though. Additionally, a theoretical study of the feasibility of mounting lidars in the blades of wind turbines for active pitch angle control has been undertaken with a positive outcome encouraging an experimental trial to measure wind with a such construction. Therefore, a small telescope CW lidar designed for turbine blade integration has been tested in a high performance wind tunnel, and very good agreement with reference measurements has been obtained.