MEASUREMENT OF THIN CLOUD OPTICAL PROPERTIES USING A COMBINED MIE-RAMAN LIDAR

Yonghua Wu,S. Chaw,B. Gross,Yu Zhao,F. Moshary

Abstract:A multiple-wavelength Mie-Raman lidar has being developed for aerosol, cloud and water vapor measurement over the Metropolitan area of New York. In this presentation, extinction, backscatter and lidar ratios profiles are obtained under daylight conditions for both aerosol and low water phase clouds. To validate the profiles, the optical depths of the low altitude clouds are derived using two independent methods. In particular, excellent agreement for the optical depth measurements between the regression technique using elastic backscatter profiles below and above the cloud and the integrated extinction from the N2-Raman returns is obtained. Furthermore, it is shown that to obtain accurate optical depths, a modified regression approach accounting for aerosol backscatter below cloud base is needed. Finally, the usefulness of the raman profiles for determining the effective water droplet size is presented. 1. MIE-RAMAN LIDAR DESCRIPTION A multiple-wavelength Mie-Raman lidar has been built for aerosol and cloud measurements at CCNY campus (40.819°N/73.949°W), New York. The transmitter is a Nd:YAG laser (Quanta-Ray PRO-230, 30Hz).operating at 355nm, 532nm and 1064nm. A Newtonian telescope with a 1.5-mrad field of view and receiver diameter 50.4-cm is used to collect the backscatter returns. Elastic returns (Mie and Rayleigh) at 355-532-1064-nm along with N2-Raman returns at 387 nm are received simultaneously. TRhe schematic layout is shown in figure 1. Fig. 1 Raman Lidar configuration Three photomultiplier tubes (Hamamatsu PMT) are used for the UV-Visible returns while a Si-APD detector is used for 1064-nm channel. A narrow-band interference filter (Barr Assoc., BW=0.3-nm) is used to suppress the skylight background noise and act as an interference filter for the N2-Raman channel. The filter is designed to effectively eliminate interference from the elastic returns. The elimination of cross-talk was confirmed by the absence of a raman signal due resulting from the strong signal due to low clouds. The detector outputs are acquired by Licel transient recorders, which combine A/D and photon counting techniques (16-bit 40 MHz A/D and 250 MHz photon counter). In our standard processing mode, Lidar signal profiles are recorded by 1-min average (1800 laser shots) and 3.75 m range resolution. For our system, the overlap function is close to unity to within a few hundred meters even though it is coaxial. 2. MEASUREMENT THEORY Aerosol and cloud retrievals by Mie-Raman lidar are well described in the literatures . Briefly, aerosol extinction can be directly derived from N2-Raman signals provided we account for the molecular extinction and backscatter (i.e. number density). To extrapolate the extinction from the raman channel at 387 nm to the elastic channel at 355 nm, an angstrom exponent is needed and taken to be 1.0 for aerosol and 0.0 for cloud. While the derived aerosol extinction could then be fed into the single scatter lidar equation for the elastic channel, it is more convenient to obtain the backscatter ratio ( ) m m a R β β β + = by taking the ratio between the elastic signal and the Raman signal. From these two independent measurements, the extinction-to-backscatter ratio, or lidar ratio, can be calculated. In these calculations, the molecular extinction and number density data are calculated using radiosonde data at Brookhaven site (OKX Upton), Southeast about 57 miles far from lidar site. Uncertainties of aerosol extinction and backscatter are mainly caused the signal noises compared to the weak Raman returns in the daytime. To assess the accuracy and consistency of these measurements, we first calculate the optical depth of low altitude clouds which are sufficiently thin so that the lidar may penetrate above the cloud with sufficient SNR. In making this calculation, the optical depth of the cloud can obviously be obtained by integrating the extinction profiles through the cloud. However, the optical depth of a thin cloud can also be derived by regressing the theoretical molecular returns to the measured elastic returns below and above cloud, respectively . The method given above is applicable only when no aerosol exists above or below cloud base and it therefore primarily used for high cirrus. However, in our case, significant aerosol will occur so the formulation must be slightly modified. To this end, we consider the correction of aerosol scattering ratio (ratio of total backscatter to only molecular backscatter) for the low clouds. Below cloud (range z1~z2) and above cloud (z3~z4), slopes are referred to, respectively: ) , ( ) , ( 2 1 1 0 2 z z R z z CT K a = − (1)