Standoff detection of backscattered light based on tunable diode laser absorption spectroscopy

Abstract

Standoff detection is of great significance in greenhouse gas measurement monitoring, early warning security and leak detection. Variations in topographic characteristics and distances will undoubtedly increase background noise. However, the effect of real-time changes in the angle between the sensor and the target on dynamic detection is rarely studied. In this article, a portable laser sensor is developed to retrieve gas concentration of CH4 by collecting the backscattered light from a distant non-cooperative target. The experimental results show that “2f/1f” based on wavelength-locked wavelength modulation technique can effectively suppress the background interference caused by a change in the detection angle. Increasing the sampling frequency does not achieve denoising for static detection but can lead to significant noise reduction for dynamic detection. For a modulation frequency of 4 k Hz, the mean standard deviation of the gas concentration is improved to 42.8 % and 7.04 % of that at 1 k Hz and 250 Hz, respectively. Kalman filtering can significantly improve the denoising effect for dynamic detection. The background concentration perturbation is reduced by a factor of 4.19 ∼ 4.71 after applying Kalman filtering. Combined with Kalman filtering, the minimum detection limit for Allan variance prediction can reach 20.5 ppb@ (0.15 sec) at a modulation frequency of 4 k Hz.