We describe an operational, self-contained, fully autonomous Raman lidar system that has been developed for unattended, around-the-clock atmospheric profiling of water vapor, aerosols, and clouds. During a 1996 three-week intensive observational period, the system operated during all periods of good weather (339 out of 504 h), including one continuous five-day period. The system is based on a dual-field-of-view design that provides excellent daytime capability without sacrificing nighttime performance. It is fully computer automated and runs unattended following a simple, brief (~5-min) start-up period. We discuss the theory and design of the system and present detailed analyses of the derivation of water-vapor profiles from the lidar measurements.

Light polarization is used in the animal kingdom for communication, navigation, and enhanced scene interpretation and also plays an important role in astronomy, remote sensing, and military applications. To date, there have been few photodetector materials demonstrated to have direct polarization sensitivity, as is usually the case in nature. Here, we report the realization of a carbon-based broadband photodetector, where the polarimetry is intrinsic to the active photodetector material. The detector is based on p-n junctions formed between two macroscopic films of single-wall carbon nanotubes. A responsivity up to ~1 V/W was observed in these devices, with a broadband spectral response spanning the visible to the mid-infrared. This responsivity is about 35 times larger than previous devices without p-n junctions. A combination of experiment and theory is used to demonstrate the photothermoelectric origin of the responsivity and to discuss the performance attributes of such devices.

[1] Raman lidar water vapor and aerosol extinction profiles acquired during the daytime over the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) site in northern Oklahoma (36.606 N, 97.50 W, 315 m) are evaluated using profiles measured by in situ and remote sensing instruments deployed during the May 2003 Aerosol Intensive Operations Period (IOP). The automated algorithms used to derive these profiles from the Raman lidar data were first modified to reduce the adverse effects associated with a general loss of sensitivity of the Raman lidar since early 2002. The Raman lidar water vapor measurements, which are calibrated to match precipitable water vapor (PWV) derived from coincident microwave radiometer (MWR) measurements were, on average, 5-10% (0.3-0.6 g/m 3 ) higher than the other measurements. Some of this difference is due to out-of-date line parameters that were subsequently updated in the MWR PWV retrievals. The Raman lidar aerosol extinction measurements were, on average, about 0.03 km À1 higher than aerosol measurements derived from airborne Sun photometer measurements of aerosol optical thickness and in situ measurements of aerosol scattering and absorption. This bias, which was about 50% of the mean aerosol extinction measured during this IOP, decreased to about 10% when aerosol extinction comparisons were restricted to aerosol extinction values larger than 0.15 km À1 . The lidar measurements of the aerosol extinction/backscatter ratio and airborne Sun photometer measurements of the aerosol optical thickness were used along with in situ measurements of the aerosol size distribution to retrieve estimates of the aerosol single scattering albedo (w o ) and the effective complex refractive index. Retrieved values of w o ranged from (0.91-0.98) and were in generally good agreement with w o derived from airborne in situ measurements of scattering and absorption. Elevated aerosol layers located between about 2.6 and 3.6 km were observed by the Raman lidar on 25 and 27 May. The airborne measurements and lidar retrievals indicated that these layers, which were likely smoke produced by Siberian forest fires, were primarily composed of relatively large particles (r eff $ 0.23 mm) and that the layers were relatively nonabsorbing (w o $ 0.96-0.98). Preliminary results show that major modifications that were made to the Raman lidar system during 2004 have dramatically improved the sensitivity in the aerosol and water vapor channels and reduced random errors in the aerosol scattering ratio and water vapor retrievals by an order of magnitude.

Water vapor mass mixing ratio profiles from NASA’s Lidar Atmospheric Sensing Experiment (LASE) system acquired during the Atmospheric Radiation Measurement (ARM)–First International Satellite Cloud Climatology Project (ISCCP) Regional Experiment (FIRE) Water Vapor Experiment (AFWEX) are used as a reference to characterize upper-troposphere water vapor (UTWV) measured by ground-based Raman lidars, radiosondes, and in situ aircraft sensors over the Department of Energy (DOE) ARM Southern Great Plains (SGP) site in northern Oklahoma. LASE was deployed from the NASA DC-8 aircraft and measured water vapor over the ARM SGP Central Facility (CF) site during seven flights between 27 November and 10 December 2000. Initially, the DOE ARM SGP Cloud and Radiation Testbed (CART) Raman lidar (CARL) UTWV profiles were about 5%–7% wetter than LASE in the upper troposphere, and the Vaisala RS80-H radiosonde profiles were about 10% drier than LASE between 8 and 12 km. Scaling the Vaisala water vapor profiles to match the precipitable water vapor (PWV) measured by the ARM SGP microwave radiometer (MWR) did not change these results significantly. By accounting for an overlap correction of the CARL water vapor profiles and by employing schemes designed to correct the Vaisala RS80-H calibration method and account for the time response of the Vaisala RS80-H water vapor sensor, the average differences between the CARL and Vaisala radiosonde upper-troposphere water vapor profiles are reduced to about 5%, which is within the ARM goal of mean differences of less than 10%. The LASE and DC-8 in situ diode laser hygrometer (DLH) UTWV measurements generally agreed to within about 3%–4%. The DC-8 in situ frost point cryogenic hygrometer and Snow White chilled-mirror measurements were drier than the LASE, Raman lidars, and corrected Vaisala RS80H measurements by about 10%–25% and 10%–15%, respectively. Sippican (formerly VIZ Manufacturing) carbon hygristor radiosondes exhibited large variabilities and poor agreement with the other measurements. PWV derived from the LASE profiles agreed to within about 3% on average with PWV derived from the ARM SGP microwave radiometer. The agreement between the LASE and MWR PWV and the LASE and CARL UTWV measurements supports the hypotheses that MWR measurements of the 22-GHz water vapor line can accurately constrain the total water vapor amount and that the CART Raman lidar, when calibrated using the MWR PWV, can provide an accurate, stable reference for characterizing upper-troposphere water vapor