Pinatubo effects in the Arctic as derived from Airborne and Surface observations

Pinatubo effects in the Arctic as derived from Airborne and Surface observations

Acknowledgments. This work was supported by ONR grant N00014-90-J-1840, NASA grant NAGW-2407, and NOAA/CMDL. Special thanks are due to the NOAA WP-P3 crew, P. Sheridan, R. Schnell, G. Herbert, and H. Bridgman for their assistance in the field and access to ancillary data.


The June 1991 eruptions of Mount Pinatubo ejected massive amounts of debris and sulfur dioxide gas into the stratosphere that were dispersed globally by upper-level winds. By late winter 1992 the spread of volcanic aerosols had reached the high latitudes reaching peak opacity during April in the western Arctic. The climatic impacts of Pinatubo were quite significant and of global proportion. Here we discuss only how the spread of Pinatubo aerosols affected the Alaskan Arctic using data collected during aircraft flights over that region and surface observations made at the NOAA/CMDL Barrow Observatory (BRW). Using these data important characteristics of the Pinatubo aerosol in the Arctic were determined, including their spectral aerosol optical depth AOD, inferred size spectra, and their rate of decay.

During the spring of 1992 an extensive series of in situ measurements were made using airborne techniques as part of the Fourth Arctic Gas and Aerosol Sampling Program (AGASP-IV) in conjunction with the Arctic Leads Experiment (LEADEX). Nearly 1300 spectral measurements of solar irradiance were made from near the surface into the stratosphere using handheld sunphotometers during seven flights of the NOAA WP-3D (P3) aircraft pictured in Figure 1. We focus here on an analysis of the stratospheric data to quantify the spectral opacity and infer effective size distributions for the Pinatubo aerosols that were present in the Arctic. Ancillary surface measurements are used to estimate a decay rate of stratospheric optical depth following the period of peak aerosol concentration. These results are discussed in greater detail in Stone et al., (1993), and in Stone et al. (1994).

Figure 1. The NOAA WP-3D.


The data are derived from two types of radiometric observations: (1) airborne measurements made using multi-channel sunphotometers, and (2) "wideband" surface-based direct beam solar irradiance measurements. The sunphotometer observations were made using handheld instruments that sense narrowband, solar irradiance at 380, 500, 778, and 862 nm. The wideband data were collected at the NOAA Climate Monitoring and Diagnostics Laboratory (CMDL) observatories. Only data collected during cloud-free periods are analyzed. The locations and dates corresponding to the various measurement periods are shown in Figure 2. The airborne observations were made through optical windows specially designed for the P3 aircraft. All optical depth values reported here account for Rayleigh scattering, ozone absorption at 500 nm, and changes in relative airmass as a function of time and location, as well as the attenuation caused by the window. From series of slant-path irradiance measurements, columnar aerosol optical depths were calculated with an accuracy of about 0.02 optical depth units ODUs. The wideband data derived from suface measurements are accurate to within ±0.04 ODUs.

Figure 2. Distribution of sunphotometer measurements made during AGASP-IV/LEADEX stratospheric flight segments and at surface locations (ANCH, TALK, and SIMMS). Wideband pyrheliometric measurements were made at the NOAA/CMDL Barrow Observatory (BRW).


3.1. Aerosol Optical Depth

Figure 3 shows an example of the spectral aerosol optical depth data that was obtained during of the AGASP-IV flights. The figure shows a time series of (color-coded) spectral AOD measured during Flight 402, which was flown over northern Alaska and in the vicinity of Pt. Barrow. The pressure altitudes during various flight segments (right scale) when sunphotometer measurements were made are also indicated. This particular flight provided an excellent opportunity to make measurements from near the surface into the stratosphere, thus enabling an assessment of the vertical distribution of aerosols throughout the atmospheric column. Figure 4 shows such a "profile" flown from Prudhoe Bay to the west of Barrow out over the Beaufort Sea. The enhanced optical depths measured near the surface are attributed to the existence of a haze layer, possibly of anthropogenic origin, transported into the region across the Arctic Ocean.

Figure 3. Time series of spectral aerosol optical depth measurements made during AGASP IV flight 402 (color coded, left scale). Pressure altitude of Aircraft is shown as slate blue trace referenced to the right scale.
Figure 4. Optical depth profile from Flight 402 flown from Prudhoe Bay to the west of Barrow out over the Beaufort Sea between approximately 12:30 and 14:30 LST. Refer to Figure 3.

By differencing successive, vertically resolved measurements of (total column) AOD, accounting for layer thickness, it is possible also to derive a profile of the volume extinction coefficient K in units of km-1, where the optical depth t = Kdz, and dz is the layer thickness. Figure 5 was constructed from the profile shown in Figure 4 to illustrate how extinction varies with height.

Figure 5. Volume extinction profile derived from Figure 4 for Flight 402 (cyan) compared with subarctic models of aerosol extiction derived from LOWTRAN 7. Actual tropopause height of flight indicated for reference.

The observed profile is compared with three model profiles of aerosol extinction derived from LOWTRAN 7, a widely used radiative transfer code. These were selected for subarctic conditions assuming "background" conditions throughout the atmosphere, and two levels of enhancement referred to as "high" and "extreme" loading by volcanic aerosols in the stratosphere. Because no measurements were made above 9.7 km during Flight 402, the extinction above this region is assumed to take the shape of the model profiles, but preserve the magnitude of the total integrated extinction above, which is due to the presence of the Pinatubo aerosol. Despite the fact that this distribution could not be corroborated, it is clear that a layer of volcanic aerosols in the stratosphere had extinction values of the same order of magnitude as those observed at mid-levels of the troposphere.

During Flight 402, the stratospheric extinction was at least an order of magnitude, perhaps 20-30 times, greater than normal even though the air flow was from within the polar vortex, which at this time did not contain high levels of Pinatubo aerosols. As the vortex began to break up, higher concentrations of volcanic aerosols were mixed into the stratosphere due to transport from the south. On subsequent AGASP-IV flights stratospheric AOD exceeded 0.2 in the visible range as indicated in Figure 6 for notes on Flight 407. Similar time series for other flights appear in Fett et al., Part I.

Figure 6. Time series of spectral aerosol optical depth measurements made during AGASP IV flight 407 (color coded, left scale). Pressure altitude of Aircraft is shown as slate blue trace referenced to the right scale.

AOD values measured during this flight were more typical of other AGASP-IV flights. There was an absence of haze and the total column AOD was dominated by volcanic aerosols in the stratosphere. Figure 7 shows the magnitude of this component compared with "background" tropospheric conditions that were determined from years of similar measurements made at BRW, the NOAA/CMDL observatory closest to where AGASP-IV flights took place. The figure shows the mean stratospheric aerosol optical depths, plus and minus one standard deviation derived from sunphotometer observations made during Flight 407 while the P3 was flying above the tropopause (see map). Not only is the magnitude of stratospheric AOD large compared with normal tropospheric values, but the spectral signature is significantly different, tending to be nearly "flat" in the visible range and falling off rather sharply for wavelength > 778 nm. The low standard deviation of measurements made in the stratosphere and the fact that theses features were typical of other flights indicate there was a high degree of homogeneity in time and space during late April 1992 in this region of the Arctic.

Figure 7. Spectral aerosol optical depths derived for a stratospheric segment during AGASP-IV Flight 407. The lower curve (adapted Dutton et al., 1984) gives the values of the tropospheric background. Curves showing plus and minus one standard deviation of the measurements are also indicated (thin gray curves).
3.2. Inferred Aerosol Size Distributions

Based on the data presented in Figure 7, effective aerosol size distributions were inferred using the inversion algorithm of King et al. (1978). A radius range, rmin = 0.15 ± 0.05 µm and rmax = 1.10 ± 0.20 µm, was prescribed, and an index of refraction of 1.45 - 0.0i assumed. The inversion results are presented in Figure 8 as curves showing the columnar number concentration dN as a function of radius, dlog(r), for the atmospheric column above the altitude where optical depth was measured in each case. The thin gray curves indicate the ranges in number density corresponding to the spectral optical depth data ± one standard deviation shown in Figure 7.

Figure 8. Effective aerosol size distributions showing the range of number concentrations dN at each radius interval [dlog(r)] inferred from the optical depth data shown in Figure 7.

As expected, in the absence of Arctic haze, the background aerosol in northern Alaska is characterized by a distribution showing a decrease in number as radius increases, often referred to as a Junge distribution and expressed by a power law function (a straight line on a log-log plot). In distinct contrast, the distribution for Flight 407 is characterized as being bimodal having moderately high concentrations of small particles, and a second, large particle peak with mode radius of about 0.5 microns. This bimodal feature of the Pinatubo aerosol was even more pronounced in size spectra inferred from other AGASP-IV AOD data, and of course evolves over time. This evolution is described in more detail in Stone et al., (1993), and has been corroborated by several independent observations made at diverse geographical regions. In every case, the Pinatubo aerosol size spectra were best represented by bimodal, lognormal distributions exhibiting a superposition of a monodisperse large-particle (volcanic) mode on a small-particle (background) mode.

3.3. Time Decay of Stratospheric Opacity

To evaluate the time decay of the Pinatubo aerosol layer(s) geographically, we analyzed the CMDL wideband optical depth records. Assuming that optical depth t decays exponentially in time t after reaching its peak value tpeak, i.e., t(t) = tpeakexp(-t/T), the time series of monthly wideband optical depth anomalies were fitted using a least-squares technique to estimate the time constants T, or e-folding times, of the aerosol layers for each of the NOAA/CMDL observatories. Only the results for BRW and MLO are discussed here. Peak optical depths occurred during August 1991 at MLO (19.5°N), but not until April 1992 at BRW (71.3°N). Stone et al. (1994) suggest that the formation of the north polar vortex during autumn 1991 prevented the penetration of the Pinatubo aerosols into that region until its breakup the following spring. It appears that the size and position of the vortex may cause local gradients in stratospheric opacity as observed by contrasting the results of Flight 402 and Flight 407 above (Figure 3, and Figure 6, respectively. Figure 9 shows the monthly mean optical depth anomalies (ODAs) determined from long-term records from BRW and MLO. Exponential fits referenced to the month of peak anomalies attributed to the Pinatubo are plotted for each site.

Figure 9. Time series of monthly mean wideband optical depth anomalies for Barrow, Alaska BRW, and Mauna Loa, Hawaii MLO highlighting the peak and subsequent decay of the Pinatubo aerosol at the respective locations. Fits were determined assuming an exponential decay rate.

The estimates of decay rate were made based on monthly mean ODAs beginning at peak opacity and through subsequent years until background conditions were again observed. The BRW decay is based on March and April data, and the MLO decay on July and August data. The decay rates are normalized for comparison in Figure 10, in which the scaling is based on a relative percentage of peak values. An estimate of the e-folding time for each fit is given in the legend in terms of months.

Figure 10. Estimates of time decay of the Pinatubo aerosol at BRW and MLO contrasting the rates in the Arctic and tropical regions. Results are normalized to 100% of peak opacity which occurred at times shown in Figure 9. e-folding times are indicated in parentheses.

This analysis updates that published in Stone et al., (1994), and gives further credence to the speculation that Pinatubo aerosols were removed from the stratosphere at a slower rate in the polar regions than in the tropics. The estimated e-folding time is 14 months at BRW compared with 10 months at MLO. The decay appears to be especially slow in the Arctic when compared with estimates that range from 8 to 10 months for other post-volcanic periods (e.g., Hofmann and Deshler, 1987; McCormick and Trepte, 1987). The prolonged lifetime of Pinatubo aerosols in the Arctic probably resulted from the unusual dynamical and microphysical processes that resulted in a concentration of aerosols within the polar vortex. During subsequent seasons, the aerosol may have persisted due to processes that favor vaporization at high altitudes and regeneration of particles at lower altitudes (Hofmann and Rosen, 1987). The formation of the vortex each winter probably prevented further elimination by natural processes because atmospheric mixing within the vortex is minimal. Although not quantified, it appears that a similar situation existed after the 1982 eruption of El Chichon (refer again to Figure 9).

4. Summary and Conclusions

A representative set of sunphotometer data collected during spring 1992 was used to quantify the opacity of the Alaskan Arctic stratosphere and to characterize the microphysical properties of volcanic aerosols transported to that region following the eruptions of Mount Pinatubo. The observed visible optical depths (0.2) exceeded any previous high-latitude measurements made in the aftermath of earlier volcanic eruptions. Airborne and surface measurements made prior to the AGASP-IV flights and later in the Canadian Arctic (not shown) suggest that spatial and/or temporal variations occurred in the aerosol layer(s), probably in response to dynamical forcings associated with the position and extent of the polar vortex. The microphysical properties of the Pinatubo aerosols can be described in terms of an effective size distribution that is bimodal, having a large-particle (volcanic) mode at 0.50 µm and a more concentrated small-particle (background) mode peaking below 0.18 µm. On the basis of our limited data, the decay rate (e-folding time) of optical depth is estimated to be about 14 months (at Barrow), which suggests a slower decay than has been observed for earlier post-volcanic periods. Because of the increased opacity of the stratosphere and the apparent longevity of the volcanic aerosols present in the Arctic, direct radiative forcing in that region and indirect climate perturbations (via feedbacks) on a much broader geographical scale may have resulted, but have not yet been studied in any detail.


Arctic Haze
Arctic Haze: a documentary
Arctic Haze: An Uninvited Spring Guest
Vertical Aerosol size distributions in remote regions
Origins of Arctic Haze
Arctic Gas and Aerosol Sampling Program (AGASP)
Herbert, G.A., et al., Analysis of meteorological conditions during AGASP-IV: March 30-April 23, 1992, NOAA Tech. Memo. ERL CMDL-5, 118 pp., NOAA/ERL/Climate Monitoring and Diagnostics Lab., Boulder, Colo., 1993.
Pinatubo (Cascades Volcano Observatory)
Pinatubo (Volcano World-University of North Dakota)
Ash from Mt. Pinatubo
Pinatubo Volcano "The Sleeping Giant Awakens"
Effects of Mount Pinatubo Eruption on Earth's Radiation Budget


Dutton, E.G., et al., Features of aerosol optical depth observed at Barrow, March 10-20, 1983, Geophys. Res. Lett., 11(5), 385-388, 1984.

Herbert, G.A., et al., Analysis of meteorological conditions during AGASP-IV: March 30-April 23, 1992, NOAA Tech. Memo. ERL CMDL-5, 118 pp., NOAA/ERL/Climate Monitoring and Diagnostics Lab., Boulder, Colo., 1993.

Hofmann, D.J., and J.M. Rosen, On the prolonged lifetime of the El Chichon sulfuric acid aerosol cloud, J. Geophys. Res., 92(D8), 9825-9830, 1987.

King, M.D., et al., Aerosol size distributions obtained by inversion of spectral optical depth measurements, J. Atmos. Sci., 35, 2153-2167, 1978.

McCormick, M.P., and C.R. Trepte, Polar stratospheric optical depth observed between 1978 and 1985, J. Geophys. Res., 92(D4), 4297-4306, 1987.

Stone, R. S., Ellsworth G. Dutton, and Jeffrey R. Key, Properties and Decay of Pinatubo Aerosols in Polar Regions Compared with Tropical Observations, 1994.

Stone, et al., 1993: Properties and decay of stratospheric aerosols in the Arctic following the 1991 eruptions of Mount Pinatubo, Geophys. Res. Lett., 20(21), 2359-2362,


Asano, S., et al., Spectral optical thickness and size distributions of the Pinatubo volcanic aerosols as estimated by ground-based sunphotometry, J. Meteorol. Soc. Jpn., 71(1), 165-173, 1993.

Bluth, G.J.S., et al., Global tracking of the SO2 clouds from the June, 1991 Mount Pinatubo eruptions, Geophys. Res. Lett., 19(2), 151-154, 1992.

Davidson, K.L., J.E. Overland and D.E. Wolfe, Leadex Data Report, Part 2, Rawinsonde and Ice Station Data, Naval Research Laboratory, NRL/PU/7541--92-0004, 561 pp., 1992.

Deshler, D., et al., Balloonborne measurements of the Pinatubo aerosol size distribution and volatility at Laramie, Wyoming during the summer of 1991, Geophys. Res. Lett., 19, 199-202, 1992.

Dutton, E.G., and J.R. Christy, Solar radiative forcing at selected locations and evidence for global lower tropospheric cooling following the eruptions of El Chichon and Pinatubo, Geophys, Res. Lett., 19(23), 2313-2316, 1992.

Dutton, E.G., et al., Shortwave aerosol optical depth of Arctic haze measured on board the NOAA WP-3D during AGASP-II, April 1986, J. Atmos. Chem., 9, 71-79, 1989.

Fett R. W., Ed., Leadex Data Report, Part 3, Aircraft Data and Flight Summaries, Naval Research Laboratory, NRL/PU/7540--93-0015, 302 pp., 1993.

Fett R. W., Ed., Leadex Data Report, Part 1, Weather Analysis and Satellite Images, Naval Research Laboratory, NRL/PU/7531--92-0003, 409 pp., 1992.

Gerber, H.E., and Deepak, A., Eds., Aerosols and Their Climatic Effects, A. Deepak Publishing, 1984.

Hansen, J., et al., Potential climate impact of Mount Pinatubo eruption, Geophys. Res. Lett., 19(2), 215-218, 1992.

Harris, J.M., and B.A. Bodhaine, Eds., Geophysical Monitoring for Climatic Change No. 11: Summary Report 1982, 160 pp., NOAA/ARL/GMCC, Boulder, Colo., 1983.

King, M.D., et al., Aerosol size distributions obtained by inversion of spectral optical depth measurements, J. Atmos. Sci., 35, 2153-2167, 1978.

Lacis, A., J. Hansen, and M. Sato, Climate forcing by stratospheric aerosols, Geophys. Res. Lett., 19(15), 1607-1610, 1992.

McCormick, M.P., et al., Spatial changes in the stratospheric aerosol associated with the north polar vortex, Geophys. Res. Lett., 10(10), 941-944, 1983.

Minnis, P., et al., Radiative climate forcing by the Mount Pinatubo eruption, Science, 259, 1411-1415, 1993.

Reddan, et al., The SIMMS'92 data report, ISTS-EOL-SIMS-TR-92-003, Earth Observations Laboratory, University of Waterloo, 226 pp., 1992.

Reddy, P.J., et al., Aerosol optical depths over the Atlantic derived from shipboard sunphotometer observations during the 1988 global change expedition, Global Biochem. Cycles., 4(3), 225-240, 1990.

Robock, A., and J. Mao, Winter warming from large volcanic eruptions, Geophys. Res. Lett., 12(24), 2405-2408, 1992.

Shaw, G.E., Sun photometry, Bull. Am. Meteorol. Soc., 64(1), 4-10, 1983.

Spinhirne, J.D., and M.D. King, Latitudinal variation of spectral optical thickness and columnar size distribution of the El Chichon stratospheric aerosol layer, J. Geophys. Res., 90(D6), 10, 607-10, 619, 1985.

Toon, O.B., and J.B. Pollack, A global average model of atmospheric aerosols for radiative transfer calculations, J. Appl. Meteorol.,15, 225-246, 1976.

Van de Hulst, H.C., Light Scattering by Small Particles, Dover Publications, N.Y., 470 pp., 1981.