3.1.5. Lidar Measurements at Mauna Loa Observatory

Monitoring of the stratospheric aerosol with ruby (694 nm) and Nd:YAG (532 nm) lidars continued as part of the activities of the MLO primary site of the Network for the Detection of Stratospheric Change (NDSC). The 2-year period of 1996 through 1997 provided an excellent opportunity to observe the stratospheric aerosol layer under background conditions. The background level of about 6 Ž 10-5 sr^-1 (at 694 nm) was reached in late 1995 and no further decay of the Mt. Pinatubo eruption was evident. Although this level was measured before in 1980 and 1990 by the ruby lidar, this is the first background period that the more accurate Nd:YAG lidar (532 and 1064 nm) has been in operation. Two striking features of the background aerosol are apparent with the YAG measurements that were not resolved by the ruby lidar system. One feature is a clear seasonal variation in aerosol backscatter seen in Figure 3.5. Here the total aerosol in the stratospheric layer is plotted which is dominated by the lower part of the layer above the tropopause. The source and sink balance of the stratospheric aerosol is affected by changes in the tropopause height. The variation is smaller than that seen at midlatitudes probably because the seasonal change in the tropopause is also much smaller.

Fig. 3.5. Integrated aerosol backscatter at 532, 694, and 1064 nm.

Figure 3.6 shows the long-term record of the ruby lidar and the record from inception in early 1994 through 1997 for the Nd:YAG. The dominant features in the data are the major volcanic eruptions of El Chichón in 1982 and Mt. Pinatubo in 1991. A comparison of these events and an analysis of the eruption-free background period during 1996 and 1997 was recently published [Barnes and Hofmann, 1997].

Fig. 3.6. Lidar backscatter at two wavelengths (red–694 nm and green–532 nm) from stratospheric aerosol between the altitudes of 15.8 and 33 km at MLO. The occurrence of volcanic eruptions which were believed to perturb the stratospheric aerosol level are indicated by arrows. The ratio of green to red backscatter decreased from 1994 to 1997 signifying a general increase in average particle size over this time period. During the background period of 1996-1997, there appears to be a QBO-related variation in the backscatter magnitude (see text).

In Figure 3.7 the aerosol backscatter above 25 km is compared with the 30 hPa tropical quasibiennial oscillation (QBO) winds; this is the second feature of the background aerosol. As the winds switch from westerly to easterly, the top of the aerosol layer changes from about 26 km to nearly 33 km. The total aerosol in the altitude range changes by a factor of 2. Although the QBO phase dominates the loading of the stratosphere at this altitude, there are episodes of transport from midlatitudes (lower aerosol loading), as in the spring of 1996, which deviate from tropical behavior.

Fig. 3.7. Hygroscopic growth factor of aerosol scattering f_RH (s_sp), measured for the marine and anthropogenically influenced cases at Sable Island.

Final results of the NDSC aerosol analysis algorithm intercomparison were presented [Steinbrecht, et. al., 1996] and the MLO analysis agreed extremely well with other analyses of identical data sets. Also a lidar temperature intercomparison was conducted during the MLO3 ozone intercomparison. An error of about 3°C above 60 km was found between the MLO lidar and the other three lidars present. Below 60 km the temperatures agreed within measurement error. The discrepancy improved to about 1°C with an improved smoothing technique above 60 km. The temperature record was recalculated with this improvement. Both results were presented by McGee et al. [1996].

Another test involved measuring three aerosol profiles at 532 nm simultaneously and was designed to quantify the instrument error of an aerosol measurement. Fifty-four shot periods (about 1.8 seconds) were alternately added to three different files for a total of 105 minutes (35 minutes per profile). The profiles were then analyzed normally. The standard deviation averaged 5.4% through the aerosol layer. This supports the value of 6% that has been used since initial tests were done 3 years ago. The original tests were complete profiles measured in succession that included natural variability. This test implies natural variability is smaller than 1% during periods of 1 or 2 hours.

A proposal was submitted to NASA in 1997 entitled, “Lidar measurements of Cirrus Clouds and Aerosols over Mauna Loa Observatory, Hawaii for CERES and SAGE-III Validation.” The proposal used preliminary measurements of cirrus backscatter correlated with long-wave IR radiation measurements. Preliminary depolarization measurements of the clouds were also used. Although the proposal was not funded, it may be submitted again since the topic is an active area of research.

[BACK] [CONTENTS] [NEXT]