Robust Calibration Method for Pure Rotational Raman Lidar Temperature Profiling in the Troposphere

Temperature is a key input parameter to many techniques in such fields as atmospheric dynamics, meteorology, climatology, and chemistry, among others. Because temperature is closely related to factors that induce climate changes, for example, the formation of clouds, it is also involved in modulating the energy budget of the earth and in determining atmospheric stability. Also, because the lower troposphere is the most closely connected part of the atmosphere to the land surface and the biosphere, it is essential to make an accurate measurement of temperature in this area in order to comprehensively understand atmospheric processes such as heat transfer and pollutant emission. The past decades witness enormous progress in the development of atmospheric temperature detection. Lidar detection has the advantage of temporal and high-range resolution detecting. Its unique capability to allow continuous monitoring of optical particle properties and temperature in the same air mass is highly beneficial. Among the lidar techniques, the pure rotational Raman (PRR) lidar has proven to be technically useful for vertical temperature profiling in the troposphere and stratosphere.

The PRR lidar technique is based on the feature that the intensities of the lines within the pure rotational Raman spectra band exhibit different dependencies on temperature. In practical applications a calibration function is required to make it possible to retrieve temperature profiles from lidar remote sensing. In practice, calibration is performed by comparison with the temperature data obtained with other approaches, such as local radiosondes. The commonly used calibration methods often underperform with low SNR because the fitting process is prone to be affected by signal noise. In reality, large calibration errors occur for some low signal-to-noise ratio (SNR) situations restricted by lidar performance or atmosphere conditions. Such errors make temperature measurement unreliable.

In this presentation, a new temperature retrieval method for PRR lidar is proposed. Considering the change in the distribution of the PRR states of the atmospheric molecules with temperature, a temperature-sensitive part of the signal intensity ratio is constructed. This part is calculable with the radiosonde temperature data. The rest of the intensity ratio mainly determined by the lidar spectrometer characteristics is calibrated, and the temperature is retrieved along with it. The proposed method is examined through simulations and experiments using the mobile PRR lidar system in the Beijing Institute of Technology. Furthermore, we notice that there are some differences in temperature sensitivity between Stocks and anti-Stocks branches of PRR spectrum. For systems that receive both sides of PRR spectrum, we introduce an extra temperature-factor to correct the calibration error caused by these differences. Compared with the traditional fitting methods, the new method is more accurate and requires less SNR. Inversion errors are effectively reduced by ~50% within the detection distance of 5 km. The proposed method can be applied to small-scale, low-cost PRR lidar systems for lower-atmosphere detecting, especially in weak SNR environments.