Abstract

The Cloud–Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) satellite was launched in April 2006 to provide global vertically resolved measurements of clouds and aerosols. Correct discrimination between clouds and aerosols observed by the lidar aboard the CALIPSO satellite is critical for accurate retrievals of cloud and aerosol optical properties and the correct interpretation of measurements. This paper reviews the theoretical basis of the CALIPSO lidar cloud and aerosol discrimination (CAD) algorithm, and describes the enhancements made to the version 2 algorithm that is used in the current data release (release 2). The paper also presents a preliminary assessment of the CAD performance based on one full day (12 August 2006) of expert manual classification and on one full month (July 2006) of the CALIOP 5-km cloud and aerosol layer products. Overall, the CAD algorithm works well in most cases. The 1-day manual verification suggests that the success rate is in the neighborhood of 90% or better. Nevertheless, several specific layer types are still misclassified with some frequency. Among these, the most prevalent are dense dust and smoke close to the source regions. The analysis of the July 2006 data showed that the misclassification of dust as cloud occurs for <1% of the total tropospheric cloud and aerosol features found. Smoke layers are misclassified less frequently than are dust layers. Optically thin clouds in the polar regions can be misclassified as aerosols. While the fraction of such misclassifications is small compared with the number of aerosol features found globally, caution should be taken when studies are performed on the aerosol in the polar regions. Modifications will be made to the CAD algorithm in future data releases, and the misclassifications encountered in the current data release are expected to be reduced greatly.

Accurate air quality forecasts can allow for mitigation of the health risks associated with high levels of air pollution. During September 2003, a team of NASA, NOAA, and EPA researchers demonstrated a prototype tool for improving fine particulate matter (PM_2.5) air quality forecasts using satellite aerosol observations. Daily forecast products were generated from a near-real-time fusion of multiple input data products, including aerosol optical depth (AOD) from the Moderate Resolution Imaging Spectroradiometer (MODIS)/Earth Observing System (EOS) instrument on the NASA Terra satellite, PM_2.5 concentration from over 300 state/local/national surface monitoring stations, meteorological fields from the NOAA/NCEP Eta Model, and fire locations from the NOAA/National Environmental Satellite, Data, and Information Service (NESDIS) Geostationary Operational Environmental Satellite (GOES) Wildfire Automated Biomass Burning Algorithm (WF_ABBA) product. The products were disseminated via a Web interface to a small group of forecasters representing state and local air management agencies and the EPA. The MODIS data improved forecaster knowledge of synoptic-scale air pollution events, particularly over oceans and in regions devoid of surface monitors. Forecast trajectories initialized in regions of high AOD offered guidance for identifying potential episodes of poor air quality. The capability of this approach was illustrated with a case study showing that aerosol resulting from wildfires in the northwestern United States and southwestern Canada is transported across the continent to influence air quality in the Great Lakes region a few days later. The timing of this demonstration was selected to help improve the accuracy of the EPA's AIRNow (www.epa.gov/airnow/) next-day PM_2.5 air quality index forecast, which began on 1 October 2003. Based on the positive response from air quality managers and forecasters, this prototype was expanded and transitioned to an operational provider during the summer of 2004.

Abstract

Descriptions are provided of the aerosol classification algorithms and the extinction-to-backscatter ratio (lidar ratio) selection schemes for the Cloud-Aerosol Lidar and Infrared Pathfinder Satellite Observations (CALIPSO) aerosol products. One year of CALIPSO level 2 version 2 data are analyzed to assess the veracity of the CALIPSO aerosol-type identification algorithm and generate vertically resolved distributions of aerosol types and their respective optical characteristics. To assess the robustness of the algorithm, the interannual variability is analyzed by using a fixed season (June–August) and aerosol type (polluted dust) over two consecutive years (2006 and 2007). The CALIPSO models define six aerosol types: clean continental, clean marine, dust, polluted continental, polluted dust, and smoke, with 532-nm (1064 nm) extinction-to-backscatter ratios S_a of 35 (30), 20 (45), 40 (55), 70 (30), 65 (30), and 70 (40) sr, respectively. This paper presents the global distributions of the CALIPSO aerosol types, the complementary distributions of integrated attenuated backscatter, and the volume depolarization ratio for each type. The aerosol-type distributions are further partitioned according to surface type (land/ocean) and detection resolution (5, 20, and 80 km) for optical and spatial context, because the optically thick layers are found most often at the smallest spatial resolution. Except for clean marine and polluted continental, all the aerosol types are found preferentially at the 80-km resolution. Nearly 80% of the smoke cases and 60% of the polluted dust cases are found over water, whereas dust and polluted continental cases are found over both land and water at comparable frequencies. Because the CALIPSO observables do not sufficiently constrain the determination of the aerosol, the surface type is used to augment the selection criteria. Distributions of the total attenuated color ratios show that the use of surface type in the typing algorithm does not result in abrupt and artificial changes in aerosol type or extinction.