In environmental gas sensing applications, sensors with low detection limit, high temporal resolution and large area coverage are much needed to resolve the gas emission sinks and sources. Both remote sensing systems and extractive sensors benefit from the high sensitivity and selectivity of laser spectroscopy. For sensors operating in the field, the laser system faces additional challenges from the uncertainty of the surroundings, such as turbulence, power fluctuations, speckles that add extra noise to the signal etc. Techniques such as heterodyne detection and phase-sensitive measurements can help resolve some of the problems, but they also complicate the signal processing process. This dissertation studies noise sources and their mitigation using heterodyne detection in chirped laser dispersion spectroscopy technique. We examine the design concerns and implementation of a multi-modality remote sensor. The advantages and challenges of high-bandwidth real time processing are explored through the algorithm development and its experimental verification for both chirped laser dispersion spectroscopy and dual comb spectroscopy.