Using stratospheric aerosols lidar measurements from Mount Pinatubo to simulate its radiative effects Empleo de medidas lidar de aerosoles estratosféricos del Monte Pinatubo para simular sus efectos radiativos Juan Carlos Antuña (*), Aramís Fonte Camagüey Lidar Station, Camagüey Meteorological Center, INSMET, Cuba. * Email: anadelia@caonao.cu Recibido / Received: 20 Jul 2007. Versión revisada / Revised version: 31 Oct 2007. Aceptado / Accepted: 19 Nov 2007. ABSTRACT: Lidar derived stratospheric aerosol extinction profiles from the Mt. Pinatubo eruption are used for simulating the volcanic cloud radiative effects over the surface at Camagüey, Cuba. Lidar backscattering measurements of the Mt Pinatubo stratospheric aerosols cloud conducted at Mauna Loa, Hawaii, (19.5 N, 155.6 W) and Camagüey, Cuba, (21.4 N, 77.9 W) are combined in a unique dataset. Both sets of lidar measurements have been validated with SAGE II (Stratospheric Aerosol and Gas Experiment II) stratospheric aerosols extinction measurements, demonstrating its consistency. Both original lidar backscattering profiles datasets were converted to extinction profiles at 0.532 µm with 500 m vertical resolution. Monthly mean aerosol optical depth values calculated for the period July 1991 to November 1993. Using the radiative transfer code of the GCM from the GFDL simulations of the Mt Pinatubo stratospheric aerosols clouds were conducted. Simulations were conducted at noon for the 15 th day of each month in the period cited before. Both non-perturbed and perturbed conditions were simulated, in the last case using the lidar derived AOD values. For comparison purposes measurements of direct radiation conducted at Camagüey has been used. Those measurements cover both the perturbed and non-perturbed periods. Comparison results show consistent agreement between theoretical and experimental values. Keywords: Lidar, Stratospheric Aerosols, Volcanic Effects. RESUMEN: Se emplean perfiles de extinción de aerosoles estratosféricos obtenidos con lidar, correspondientes a la erupción del Monte Pinatubo, para simular los efectos radiativos de la nube volcánica sobre la superficie en Camagüey, Cuba. Los datos de las observaciones de retrodispersión por los aerosoles estratosféricos del Monte Pinatubo, realizadas con lidar en Mauna Loa, Hawaii, (19.5 N, 155.6 W) y Camagüey, Cuba, (21.4 N, 77.9 W) se combinan en una única serie. Ambas series de datos han sido validadas con las mediciones de extinción de aerosoles estratosféricos tomadas por el SAGE II (Experimento Estratosférico de Gases y Aerosoles II), demostrando su consistencia. Los perfiles originales de retrodispersión por aerosoles de ambas series de fueron convertidos a perfiles de extinción a la longitud de onda de 0.532 µm con una resolución vertical de 500 m. Se calcularon los valores medios mensuales de espesor óptico para el periodo Julio de 1991a Noviembre del 1993. Empleando el código radiativo del Modelo de Circulación Global del GFDL se realizaron simulaciones del efecto radiativo de la nube de aerosoles estratosféricos del Monte Pinatubo. Las simulaciones se realizaron para el 15 to día de cada mes en el periodo antes mencionado. Fueron simuladas las condiciones no perturbadas y las perturbadas, en el último caso usando los valores de espesor óptico obtenidos con lidar. Se emplearon observaciones de radiación solar en superficie realizadas en Camagüey con el fin de compararlas con los resultados de las simulaciones. Estas mediciones abarcan tanto el periodo perturbado como el no perturbado. Los resultados de la comparación muestran coincidencias consistentes entre los valores simulados y los experimentales. Palabras clave: Lidar, Aerosoles Estratosféricos, Efectos Volcánicos. Opt. Pura Apl. 41 (2) 159-163 (2008) - 159 - Sociedad Española de Óptica
. REFERENCES AND LINKS [1] A. Robock, Volcanic eruptions and climate, Rev. Geophys. 38, 191-219 (2000). [2] G. Wendler, Y. Kodama, Some remarks on the stratospheric temperature for Fairbanks, Alaska, after the El Chichon Eruption, Atmos. Ocean 24, 346-352 (1986). [3] G. L. Stenchikov, I. Kirchner, A. Robock, H.-F. Graf, J. C. Antuña, R. G. Grainger, A. Lambert, L. Thomason, Radiative forcing from the 1991 Mount Pinatubo volcanic eruption, J. Geophys. Res. 103, 13837-13857 (1998). [4] S. Ramachandran, V. Ramaswamy, G. L. Stenchikov, A. Robock, Radiative impacts of the Mt. Pinatubo volcanic eruption: Lower stratospheric response, J. Geophys. Res. 105, 24409-24429 (2000). [5] J. C. Antuña, Lidar measurements of stratospheric aerosols from Mount Pinatubo at Camagüey, Cuba, after some volcanic eruptions, Atmos. Environ. 30, 1857-1860 (1996). [6] J. E. Barnes, D. J. Hofmann, Lidar measurements of stratospheric aerosol over Mauna Loa Observatory, Geophys. Res. Lett. 24, 1923-1926 (1997). [7] S. M. Freidenreich, V. Ramaswamy, A new multiple-band solar radiative parameterization for general circulation models, J. Geophys. Res. 104, 31389 31409 (1999). [8] S. M. Freidenreich, V. Ramaswamy, Refinement of the geophysical fluid dynamics laboratory solar benchmark computations and an improved parameterization for climate models, J. Geophys. Res. 110, D17105, 43-50 (2005). [9] J. C. Antuña, A. Fonte, R. Estevan, B. Barja, R. Acea, J. C. Antuña Jr., Solar radiation data rescue at Camagüey, Cuba, submitted to Bull. Amer. Meteorol. Soc. (2007) [10] J. C. Antuña, I. Pomares, R. Estevan, Temperature trends at Camagüey, Cuba, after some volcanic eruptions. Atmósfera 9, 241-250 (1996). [11] R. Estevan, J. C. Antuña, Updated Camagüey lidar dataset: validation with SAGE II, Opt. Pura Apl. 39, 85-90 (2006). [12] G. B. Barjas, PhD Dissertation, in preparation (2007). [13] M. Sutter, B. Dürr, R. Philipona, Comparison of two radiation algorithms for surface-based cloud-free sky detection, J. Geophys. Res. 109, D17202, 1-7 (2004) [14] UOSMLAB, Pacific Northwest Solar Radiation Data. Physics Department-Solar Energy Center. University of Oregon Solar Monitoring Laboratory (1999). [15] R. E. Bird, and R. L. Hulstron, Clear sky broadband solar radiation model, SERI Technical Report, SERI/TR-642-761 (1991). 1. Introduction Climate effects of the most intense volcanic eruptions injecting big amounts of aerosols in the stratosphere have been the subject of numerous studies. The most notorious effect is the cooling of the earth s surface in the period of one and three years after the eruption. Other effects are the stratospheric heating, winter warming on the Northern Hemisphere continental areas, ozone depletion and cirrus cloud seeding [1]. The main mechanism of impact of the stratospheric aerosols over climate is the radiative forcing. At the surface it is characterized by the decrease of the direct solar radiation and the increase of the diffuse radiation [2]. Modeling of those climatic effects is an important contribution to the current effort to understand the physical mechanisms predominating at the different spatio-temporal scales. Those studies also contribute to the development of forecasting capabilities in case of future volcanic eruptions. Noticeable progresses have been reached in the modeling of global scale effects [3, 4]. But up to the present few studies focus on modeling local effects and less in the tropical region. Here we present the results of the simulation of the radiative effect in the surface in Camagüey, Cuba. We use lidar measurements of the Mt. Pinatubo stratospheric aerosols conducted at Camagüey, Cuba [5] and Mauna Loa, Hawaii [6]. Monthly mean values of Aerosol Optical Depth (AOD) were calculated and used to feed the radiative transfer code of the GFDL GCM [7,8]. Simulations were also conducted for average non Pinatubo conditions. Results of the simulations were compared with broadband solar radiation measurements conducted at Camagüey [9]. Opt. Pura Apl. 41 (2) 159-163 (2008) - 160 - Sociedad Española de Óptica
2. Materials and methods Lidar backscattering measurements of the Mt. Pinatubo stratospheric aerosols cloud conducted at Mauna Loa, Hawaii (19.5 N, 155.6 W) and Camagüey, Cuba (21.4 N, 77.9 W) are combined in a unique data set for the period July 1991 to November 1993 [10]. Vertical profiles of backscattering values were converted to extinction profiles at the wavelength of 0.532 µm with vertical resolution of 500 m [11]. Using these extinction profiles the AOD was calculated for each day of the above mentioned period. Finally monthly mean AOD values were calculated. For simulating the effect of the stratospheric aerosols cloud on the surface irradiance we used a state of the art radiative code. It consisted of the column radiative transfer section of the GFDL GCM model [7,8]. The model was adapted to local mesoscale conditions taking into account the vertical profiles of temperature and water vapor at Camagüey [12]. Also a surface albedo of 0.2 was used in the simulations considering local estimates of that parameter. Simulations were carried out at noon for the 15 th day of each month in the period of study. For that purpose the cosine of the local sun altitude for the 15 th day of each month was derived. Model was run under perturbed conditions produced by the stratospheric aerosol cloud from Mt. Pinatubo eruption. AOD monthly mean values from lidar measurements for July 1991 to December 1993 were introduced in the model. Runs were conducted combining the AOD values with the corresponding cosine of the local sun altitude for the 15 th day of each individual month. Also a set of runs, under non-perturbed conditions, were conducted to produce monthly mean climatological values. The data set of surface irradiance measurements from Camagüey Meteorological Station was used for model validation purposes. The data set consists of manual actinometrical measurements conducted for the period 1985 to 2004. Its time resolution is 1 hour from sunrise to sunset hours. This is a dataset currently under development as part of the Camagüey radiation database with an extension of more than 30 years [9]. Such dataset covers both the pre and post Pinatubo period as well as the period of the Pinatubo s influence from July/1991 to November/1993. Both direct beam radiation (S ) and diffuse (D) radiation measurements at noon (12:00 LT) were used for the purposes of the present paper. The data set corresponding to the non-perturbed conditions period (excluding July/1991 to December/1993) showed several limitations. On top of the common data gaps, the selection of data under the classical clear sky criteria (cloudiness below 3/10) became complicated. Table I shows the cumulative percentage of S measurements conducted under all the cloudiness conditions. For D the cumulative percentage (not shown) present a similar pattern. The very low amounts of available broadband solar radiation under clear sky conditions have been pointed out recently. In particular it was demonstrated for the station of Bermudas (32.2ºN, 64.3ºW) with only 0.2% of the clear sky, total cloud cover 0 octaves, and 3.8% of the relaxed clear sky, total cloud cover 0 or 1 octaves [13]. Those limitations with the data set were more critical for the data set corresponding to perturbed conditions period from July/1991 to December/1993. We tested some of classic methods like interpolation from adjoining values (monthly values at other time), and using modelled or climatologic values, but all of them distort logical expected values, so the only possible solution was to relax the clear sky criteria to cloudiness equal or lower than 4/10. Still for the perturbed period we had to relax the clear sky criteria to 5/10 for filling data gaps on June and October 1992 and May 1993 as well as to 6/10 for the same purpose on September 1992 and October 1993. TABLE I Cumulative percentage of cases under the different cloudiness (N) conditions for the 14 unperturbed years between 1985 and 2004. N=0 N=2 N=4 N=6 N=8 N=10 J 1.1 8.2 22.8 62.7 89.4 100 F 0.6 7.3 21.8 68.3 91.2 100 M 1.5 4.1 14.7 55.7 89.4 100 A 0.8 4.9 14.8 58.1 89.3 100 M 0.6 1.5 10.3 48.6 83.7 100 J 0.4 1.1 8.4 51.9 90.5 100 J 0.3 1.2 7.6 47.1 83.2 100 A 0.3 1.3 9.4 51.5 86.3 100 S 0 0.7 7 44.7 83.8 100 O 0 1.2 9.6 51.8 85.5 100 N 1.2 3.1 11.1 50.3 85.2 100 D 0.6 1.5 13.3 53.2 85.2 100 Selected values of S and D considering the relaxed clear sky conditions were used for deriving monthly mean values for the period of 14 nonperturbed years. Clearly the extent of the dataset does not allow deriving climatological values. Although it has been pointed out that around 15 years of data are representative of the variability of the radiation parameters showing patterns and trends [14]. Those 14 year mean values could be used confidently to characterize the anomalies produced by an extreme radiation event as the one produced Opt. Pura Apl. 41 (2) 159-163 (2008) - 161 - Sociedad Española de Óptica
by the Mt. Pinatubo stratospheric aerosols cloud. In the case of the perturbed period from July 1991 to December 1993 monthly mean values of S and D, at noon, were calculated for each one of the individual months. Transmittance values (T) were calculated as the ratio between the direct normal beam radiation and the extraterrestrial solar radiation. Extraterrestrial solar radiation at the location of Camagüey was derived using a simple radiative model [15]. Anomalies values for both S and D were calculated as the difference between the mean value of S and D for each one of the months of the perturbed period and the corresponding S and D mean value for the same month in the 14 unperturbed years. 3. Results and discussion Figure 1 shows the course of T both for climatological and post-pinatubo eruption periods. In general the climatological period shows notable variability. In the case of the post-pinatubo period it shows lower values of T than the climatological values, as it was expected, because the presence of the stratospheric aerosols clouds. Only very few values of T for the post-pinatubo period are equal of higher than the climatological period, but they belong to the year 1993. That is more than one and half year after the eruption, enough time for the decrease of the stratospheric AOD in more than 2/3 with respect to the values immediately after the eruption. The decrease of T for the post-pinatubo period is more noticeable during the winters 1991-1992 and 1992-1993 and also in the summer of 1992. The figure also shows the return of T to the climatological values by the beginning of 1993. Fig. 1: Transmittance values both for climatological and for post Pinatubo periods In Fig. 2 both the simulated and measured anomalies of S for the post-pinatubo period are depicted. Here is also evident the high level of variability in the measurements of S. That is produced by the variability of the local atmospheric constituents under clear sky conditions (water vapor, tropospheric aerosols, etc.) plus the variability of S produced by the range of cloudiness we used because of the relaxed clear sky criteria. Simulated and measured S anomalies are all negatives for the complete period studied, except for February and June 1993, for the measured S anomalies. In that sense both simulations and measurements show the radiative effect produced by the Pinatubo stratospheric at Camagüey. Although, the high variability of the measurements, enhanced by the relaxed clear criteria used to overcome the lack of data, do not matches the smooth pattern of the simulated S anomalies. Linear fits of both variables are plotted to visualize the recovery trends. Both trends are positive, with values of 5.0 Wm -2 per month for simulated S anomalies and 2.3 Wm -2 per month for the measured S anomalies. In addition the statistical significance of the differences between S measurements for perturbed and non-perturbed period was tested using the T-student test. Test results showed statistical significant differences between both periods at a significance level of α=0.05. Simulated and measured anomalies of D are shown in Fig. 3. In this case also it is evident the high variability of the measured D anomalies, because of the relaxed clear sky criteria. Simulated D anomalies are positive for the entire period, while the measured D anomalies are positives until February 1993. From that month until the end of the period of study the measured D anomalies are negative in general. As in figure 1 linear fits of both variables are plotted. Both trends are negative with magnitudes of 1.4 Wm -2 per month for simulated D anomalies and 4.1 Wm -2 per month for the measured D anomalies. The former results demonstrate both from measurements and simulations the impact of the Pinatubo stratospheric aerosols over the D and S at Camagüey, Cuba. It provides for the first time magnitudes of the D and S anomalies and for their recovery trend. The values of S, both in the measurements and simulations, decreased immediately after the eruption because of the scattering and absorption by the stratospheric aerosols. The instantaneous magnitudes of such decrease are in the order of 100 to 200 Wm -2. The recovery trend for the two and half years after the Opt. Pura Apl. 41 (2) 159-163 (2008) - 162 - Sociedad Española de Óptica
eruption show values of several Wm -2 per month. For the case of D, the positive anomalies are the result of the increased scattered radiation, and its magnitudes in the order of 50 to 100 Wm -2. The capability of the model to simulate the radiative effects of the volcanic cloud will allow future studies for a set of more intense volcanic eruptions. 4. Conclussions We have demonstrated the radiative effect of the stratospheric aerosols produced by the Pinatubo eruption at the surface art Camagüey, Cuba. The magnitudes of initial the anomalies of the S and D at the surface produced by the stratospheric aerosols clouds range between 100 to 200 Wm -2 and 50 to 100 Wm -2 respectively. Also the recovery trends were determined, with positive values for S and negative values for D. The absolute order of magnitudes of both trends is several Wm -2 per month. We have validated the simulations conducted using the GFDL radiative code adjusted for the local conditions of Camagüey. It will be a valuable tool in simulating the radiative effects over Camagüey of future and past volcanic eruptions producing bigger amounts of stratospheric aerosols. Fig. 2. Monthly anomalies for simulated and measured Direct Normal Beam Radiation (S ). Also trends both for the simulated and measured anomalies are shown. Acknowledgements Authors want to express it deep gratitude to Prof. V. Ramaswamy and S. M. Freidenreich for providing the radiative code and answering many questions during the code implementation for PC. This work has been supported by the Cuban National Climate Change Research Program grant 01303177. Fig. 3: Monthly anomalies for simulated and measured Diffuse Radiation (D). Also trends both for the simulated and measured anomalies are shown. Opt. Pura Apl. 41 (2) 159-163 (2008) - 163 - Sociedad Española de Óptica