Lidar calibration at 1064-nm channel using the water-phase and cirrus clouds

Lidar calibration is important for the elastic-scattering lidar to quantify the aerosol backscatter profile. The conventional calibration approach is to normalize the lidar return to molecular reference value in the upper troposphere or stratosphere. This approach is widely used at the visible-ultraviolet wavelengths where the molecular backscatter is the dominant contribution, but it is difficult to apply to the 1064-nm channel because of the weak molecular scattering. Alternatively, calibration methods using horizontal or slant measurements require atmospheric homogeneity over relevant horizontal scales or an independently measured scattering coefficient, and are not suited to a vertically-pointing lidar. As an alternative to these approaches, clouds are good natural targets for the lidar calibration because they occur frequently and their optical properties are often quite regular. In this paper, we explored two independent calibration methods for 1064-nm channel based on a multi-wavelength elastic-Raman lidar. The first uses low-altitude water clouds which have a stable lidar ratio and the second uses high cirrus together with the sunphotometer-measured aerosol optical depth (AOD). The use of a Raman lidar is crucial to this investigation since the Raman derived extinction profile allows us to estimate cloud multiple-scattering and aerosol attenuation below the cloud using the low-level water cloud. By choosing very special atmospheric situations where both clouds are present as well as clear sky patches, we were able to assess the consistency and variability of the calibration. Strong consistency was illustrated by noting that the relative differences of their daily averages were smaller than 15%. However, the calibration constant from the low water cloud method is generally smaller and has larger fluctuation than those obtained from the high cirrus cloud. Further consistency was also indicated by generating the dual (532- and 1064 nm) aerosol backscatter under clear sky conditions. We constrain lidar integrated backscatter coefficient with sunphotometer-measured AOD to obtain the lidar ratio at 1064-nm. The results show good agreement with the ones simulated from the sunphotometer inversion data which implies the reasonable calibration constants. Most importantly however, calibration constants taken using both methods over different periods showed the compatibility of the methods as evidenced by the stability (< 7 %) over a 2-month long period in the year 2008 where the optical arrangement of the system was fixed and normalization for laser power and neutral density filter transmittance. In the long-term with the degradation of lidar optical efficiency, the calibration constants show the reasonable decrease. Therefore, if properly performed and with the necessary ancillary data from both Sunphotometer and Raman Lidar, both methods can be used seamlessly providing the opportunity to calibrate the system with much higher frequency.