Previous: Results Next: Conclusions Up: Ext. Abst.

Discussion

The OHP data set offers a unique perspective on cirrus clouds at northern midlatitudes. Unlike campaign data or satellite observations, the lidar measurements presented here were systematically taken over the period of three years, with excellent altitudinal and temporal resolution. Analysis has revealed many optically thin and SVC events, with the latter composing 23 % of the mean optical thicknesses.

Certain considerations must be taken into account when analyzing the OHP lidar data. With the ground-based lidar technique it is not possible to make measurements when there are opaque/precipitating clouds. Consequently it is possible that our statistics may be biased, especially the number of cirrus occurrences in the optical depth range of 0.3 - 1.0, category F in Figure 4. Generally at OHP there are only  50 nights where such conditions preclude the collection of data.

In our analysis we have made several reasonable assumptions. In the calculation of the optical thickness, we have assumed  of 18.2 sr from Platt and Dilley [1984]. This value was taken from lidar measurements (1978 - 1980) in the southern hemisphere and is estimated to have an error of 20 %; thus, this is the lower limit for the error of our calculated optical thickness. Another possible source of uncertainty is the temperatures measurements. Besides possible systematic errors of the sondes, there is the distance ( 120 km) between Nîmes and OHP. Two sondes are launched per day from Nîmes, one at noon and another at midnight (universal time). For this work we used the latter sonde, which typically temporally coincided with the lidar measurements. The temperature determinations of the Nîmes' sondes are more accurate than the Raman lidar technique [Hauchecorne et al., 1992]. For the optical thickness measurements, the sondes were used to determine the - 25 ° C threshold. If any part of the cloud layers was located at or above the corresponding - 25 ° C altitude, then the entire cloud was included in the calculation of the optical thickness. Hence, any errors in the temperature measurements would only slightly affect the optical thickness determinations. Finally, there have been differing ice threshold temperatures quoted for midlatitude cirrus [Heymsfield and Platt, 1984,Roscow et al., 1996,Riedi et al., 1999]. We conducted a reanalysis of our data using a threshold temperature of - 40 ° C. This change had a minimal effect on our results (e.g., an augmentation of only 2 % for SVC occurrences).