Comparison of CALIPSO aerosol optical depth retrievals to AERONET measurements, and a climatology for the lidar ratio of dust
- 1NASA Langley Research Center, Hampton, Virginia, USA
- 2Laboratoire d'Optique Atmosphérique, Université de Lillé 1, CNRS, Villeneuve d'Ascq, France
Abstract. We compared CALIPSO column aerosol optical depths at 0.532 μm to measurements at 147 AERONET sites, synchronized to within 30 min of satellite overpass times during a 3-yr period. We found 677 suitable overpasses, and a CALIPSO bias of −13% relative to AERONET for the entire data set; the corresponding absolute bias is −0.029, and the standard deviation of the mean (SDOM) is 0.014. Consequently, the null hypothesis is rejected at the 97% confidence level, indicating a statistically significant difference between the datasets. However, if we omit CALIPSO columns that contain dust from our analysis, the relative and absolute biases are reduced to −3% and −0.005 with a standard error of 0.016 for 449 overpasses, and the statistical confidence level for the null hypothesis rejection is reduced to 27%. We also analyzed the results according to the six CALIPSO aerosol subtypes and found relative and absolute biases of −29% and −0.1 for atmospheric columns that contain the dust subtype exclusively, but with a relatively high correlation coefficient of R = 0.58; this indicates the possibility that the assumed lidar ratio (40 sr) for the CALIPSO dust retrievals is too low. Hence, we used the AERONET size distributions, refractive indices, percent spheres, and forward optics code for spheres and spheroids to compute a lidar ratio climatology for AERONET sites located in the dust belt. The highest lidar ratios of our analysis occur in the non-Sahel regions of Northern Africa, where the median lidar ratio at 0.532 μm is 55.4 sr for 229 retrievals. Lidar ratios are somewhat lower in the African Sahel (49.7 sr for 929 retrievals), the Middle East (42.6 sr for 489 retrievals), and Kanpur, India (43.8 sr for 67 retrievals). We attribute this regional variability in the lidar ratio to the regional variability of the real refractive index of dust, as these two parameters are highly anti-correlated (correlation coefficients range from −0.51 to −0.85 for the various regions). The AERONET refractive index variability is consistent with the variability of illite concentration in dust across the dust belt.