Eos, Vol. 87, No. 4, 24 January 2006 Geophysics program and the NOAA Ocean Ex- ploration Program.We thank the captain and crew of the research vessel T. G. Thompson and the University of Washington for rapidly facili- tating the 2005 response effort.We also thank D. Butterfield, D. Fornari,T. Garfield, B. Glazer, S. Giovanoni, J. Lupton, D. Kadko, M. Lilley, and the entire Endeavour 2005 Response Team. The SOSUS project is supported by the NOAA Vents Program and the expertise of M. Fowler, C. Fox, J. Haxel, J. Klay, A. Lau, H. Matsumoto, and S. Merle. JPC gratefully acknowledges funding from NSF (OCE-0222069) and NASA (Univer- sity of Hawaii NASA Astrobiology Institute). This is PMEL contribution 2847.

References

Baker, E.T., G. J. Massoth, and R.A. Feely (1987), Cataclysmic hydrothermal venting on the Juan de Fuca Ridge, Nature, 329, 149–151. Bohnenstiehl, D. R., R. P. Dziak, M.Tolstoy, C. G. Fox, and M. Fowler (2004),Temporal and spatial history of the 1999–2000 Endeavour Segment seismic series, Juan de Fuca Ridge, Geochem. Geophys. Geosyst., 5, Q09003, doi:10.1029/2004GC000735. Davis, E. E., et al. (2004), Hydrological response to a seafloor spreading episode on the Juan de Fuca Ridge, Nature, 430, 335–338. Dziak, R. P. , C. G. Fox, and A. E. Schreiner (1995),The June–July 1993 seismoacoustic event at CoAxial Fig. 2. Diagram of earthquake migration rate (meters per second) versus onset of migration segment, Juan de Fuca Ridge: Evidence for a (hours after beginning of swarm) for swarms shown in Figure 1. Swarms associated with sea- lateral dike injection, Geophys. Res. Lett., 22 (2), 135–138. floor eruptions, fluid temperature changes, or megaplume events are shown as stars; swarms with Embley, R.W.,W. W. Chadwick Jr., I. R. Jonasson, D.A. no clear magmatic activity or megaplumes are shown as circles.The logarithmic decay curve illus- Butterfield, and E.T. Baker (1995), Initial results trates apparent nonlinear relationship.The subaerial 1979 Krafla (Iceland) dike injection event is of the rapid response to the 1993 CoAxial event: added for comparison. Relationships between hydrothermal and volcanic processes, Geophys. Res. Lett., 22 (2), 143–146. Palmer, M. R., and G. G. J. Ernst (1998), Generation of J. Cowen, Department of Oceanography, University of Holden, J. F. , M. Summit, and J. A. Baross (1998), hydrothermal megaplumes by cooling of pillow Hawaii; E. Baker and R. Embley NOAA/Pacifi c Marine Thermophilic and hyperthermophilic microorgan- basalts at mid-ocean ridges, Nature, 393, 643–647. isms in 3–30°C hydrothermal fluids following a Environmental Laboratory; D. Bohnenstiehl, Depart- deep-sea volcanic eruption, FEMS Microbiol. Ecol., Author Information ment of Earth and Environmental Sciences, Columbia 25, 33–41. University and Lamont-Doherty Earth Observatory; Lowell, R. P. , and L. N. Germanovich (1995), Dike injection and the formation of megaplumes at R. Dziak and B. Chadwick, Oregon State University/ and J. Resing, University of Washington/JISAO NOAA/ ocean ridges, Science, 267, 1804–1807. CIMRS NOAA/PMEL; E-mail: [email protected]; PMEL

high clouds have increased to a larger extent, leading to both an increase in cloud amount Can Earth’s Albedo and Surface (higher albedo) and an increased trapping of infrared radiation by clouds (increased Temperatures Increase Together? heating). PAGES 37, 43 ized in climate models.Thus, the scale of these variations presents a fundamental, and as yet Measuring the Earth’s Albedo Changes in climate depend essentially on unmet, challenge to understanding and pre- three basic parameters, the amount of inci- dicting the Earth’s climate. To derive ideal estimates of the Earth’s dent sunlight, the fraction of this sunlight that The sign and magnitude of albedo changes reflectance, it would be necessary to ob- is reflected by the Earth, and the trapping of over the past five years have been under de- serve reflected radiances at all angles from the Earth’s infrared radiation by greenhouse bate.The evidence from a variety of sources all points on the Earth, which is technically gases.The Earth’s reflectance of the Sun’s ra- suggests that the Earth’s albedo has increased impossible.Therefore, all measurements from diation back to space—or albedo—is the least from 2000 to 2004. However, the rising plan- which albedo can be inferred require as- well studied of the three. etary albedo, and the increase of sunlight be- sumptions and/or modeling. During recent The albedo depends primarily on cloud ing reflected back into space, has not led to a decades, there have been some efforts to properties, and ground-based and satellite reversal of global warming. measure the Earth’s albedo from space; but studies published within the past 3–4 years The most up-to-date cloud data, released in a long-term data series of the Earth’s albedo have shown a surprisingly signifi cant interan- August 2005 from the International Satellite is difficult to obtain due to the complicated nual and decadal variability in this parameter. Cloud Climatology Project (ISCCP), a careful intercalibration of the different satellite data The variability in reflectance is tied to changes compilation of cloud observations covering and the long gaps in the series. However, the in cloud location, amount, and thickness. the entire Earth from a range of meteorologi- availability of different albedo (and cloud) However, clouds are very poorly parameter- cal satellites, reveal that the explanation of databases, and their intercomparisons, can this seeming anomaly lies primarily in a redis- help to constrain the assumptions necessary B Y E. PALLÉ, P. R. GOODE, P. MONTAÑÉS- tribution of the clouds.Whereas low clouds to derive estimates.Thus, long-term ground- R ODRIGUEZ, AND S. E. KOONIN have decreased during the most recent years, based estimates of the Earth’s refl ectance, Eos, Vol. 87, No. 4, 24 January 2006 complementary to those from satellites, are an advantage. Data compiled from ground-based radiom- eter networks and sunshine recorders show, with some confidence, that sunlight reaching the ground decreased strongly (so-called global dimming) from the 1960s through the mid-1980s [Stanhill and Cohen, 2001; Wild et al., 2005].With greater confi - dence, data from the Earth Radiation Budget Experiment (ERBE, a set of satellite instru- ments designed to measure the Earth’s ener- getic balance) and Earthshine (ES, ground- Fig. 2. Globally averaged reconstruction (black) based measurements of reflectance based on of albedo anomalies from ISCCP cloud amount, the dayside earthlight reflected from the Moon optical thickness, and surface reflectance back to the nighttime observer, see Pallé et al. (following Pallé et al., 2004).The observed [2003] for details), show that more sunlight Earthshine albedo anomalies are in blue.All has been reaching the Earth’s surface from the observations agree with the reconstruction to mid-1980s to 2000 [Wielicki et al., 2002; Pallé within the 1 σ uncertainties, except for the year et al., 2004; Wild et al., 2005], although the with sparse ES data, 2003.The shaded region magnitude of these changes is still in dispute Fig. 1. (top) Globally averaged monthly mean 1999 through mid-2001 was used (as in Pallé [ Pallé et al., 2005]. total cloud amount from the ISCCP data. et al., 2004) to calibrate the reconstruction and Since 2000, ES observations indicate an The overall decrease in cloud amount from is the reference against which anomalies are increasing albedo [Pallé et al., 2004], whereas 1985 to 2000 is about 4–5% with a recovery defined.The vertical red bar indicates the esti- Clouds and the Earth’s Radiant Energy System of about 2–3% from 2000 to 2004. (bottom) mated size of the forcing by greenhouse gasses (CERES) satellite data report the opposite Globally averaged 5-year mean low (blue) since 1850. and middle + high (red) cloud amounts. result [Wielicki et al., 2005].A recent intercom- The difference in percent between low and parison of several albedo-related data sets Note that these changes in refl ectance middle + high cloud amounts is also given on strengthens the case for an increasing global imply climatologically significant changes (a top of each of the four 5-year intervals. Note few watts per square meter) in the shortwave albedo post-2000, consistent with the original the near doubling of this difference over the ES result [Pallé et al., 2005]. radiative forcing. ISCCP cloud data can also be 2000–2004 period with respect to the previ- used as inputs for global albedo models using However, it has been argued that an increas- ous means. ing albedo over the past five years is incon- ERBE bidirectional reflectance tables, which sistent with observed behavior in the global provide empirical determinations of local al- land and sea surface temperatures and ocean The difference in cloud amount between mid- bedos for various terrestrial scenes as a func- heat content [Wielicki et al., 2005]. Simply put, dle- and high-lying clouds, and low-lying clouds tion of angle of incidence and refl ection (see other climate parameters being constant, an remained around 7–8% from 1985 to 1999, but Pallé et al. [2003] and references therein for increase in the albedo implies a decrease in almost doubled (13%) over the past fi veyears details on these models).The whole-Earth al- the sunlight absorbed by the planet, thereby (see bottom panel of Figure 1).This shift toward bedos that were derived from these post-2000 leading to cooler temperatures. high cloud types decreased the negative cloud ISCCP data show the same increasing trend Clouds, though, have two opposing effects on forcing, and so increased the net. as the ES observations, and no trend at all if the Earth’s radiation budget:They refl ect short- In this way, the recent increase in albedo, the cloud amount is artifi cially fixed to that of wave radiation (a negative forcing cooling the following the increase in cloud amount, is year 2000 (not shown). planet) and they trap infrared/heat radiation (a not necessarily inconsistent with the ob- The 2005 update of the ISCCP data con- positive forcing).The net cloud radiative forcing served global temperatures. It seems that firms an increasing global albedo since 2000, is about -13 watts per square meter [Ramana- under the right circumstances either an but this is only the most recent change in than et al., 1989], but various cloud types con- increase or a decrease in albedo can lead climatologically signifi cant refl ectance varia- tribute differently to this forcing: Optically thick to higher temperatures.This explanation tions extending over the past two decades. low-lying clouds have a strong cooling effect, presents the most consistent picture for all of Accounting for such variations in global while high thin cirrus warm. the data sets. Furthermore, it is a cautionary climate models is essential to understanding note against inferring changes in net radia- and predicting climate change. The Role of Clouds tive forcing from shortwave data alone (e.g., the global dimming and subsequent global Acknowledgments In August 2005, ISCCP global cloud data brightening seen in ground-based data). were released covering 2001–2004, and this Pallé et al. [2004] used a multiple regression This research was supported in part by a most up-to-date set serves to clarify the evolu- of 1999–2001 ISCCP data with overlapping grant from NASA (NAG5-11007).The cloud tion of the albedo.The data show that the ES observations to construct from the former ISCCP D1 data sets were obtained from the cloud amount increased by 2–3% from 2000 a proxy for the Earth’s reflectance. Here that NASA Langley Research Center Atmospheric to 2004. In particular, the ISCCP cloud amount proxy has been projected forward beyond Sciences Data Center.We thank Lou Lanzerotti data show a sinusoidal behavior over the last 2001 through 2004 using the new ISCCP data. for suggesting helpful improvements to this 20 years (see top panel in Figure 1), with a The result, plotted in Figure 2, shows an in- manuscript. decline in all cloud types from the late 1980s creasing trend after 2000, in agreement with through the late 1990s; the total then began the ES data. References increasing in about 2000. The disagreement in 2003 may be associ- However, low clouds continued to decrease ated with the sparse ES data available for that Pallé, E., P. R. Goode,V.Yurchyshyn, J. Qiu, J. Hickey, post-2000, while middle and high clouds in- year: slightly more than one half of the annual P. Montañés-Rodriguez, M.-C. Chu, E. Kolbe, C.T. creased. In the bottom panel of Figure 1, the average number of observations from 1999 to Brown, and S. E. Koonin (2003), Earthshine and globally averaged cloud amount has been 2004, but with a normal seasonal distribution the Earth’s albedo: 2. Observations and simula- tions over 3 years, J. Geophys. Res., 108(D22), 4710, divided into low (p > 680 mbar) and middle (E. Pallé et al., Seasonal and interannual trends doi:10.1029/2003JD003611. combined with high (p < 680 mbar) cloud in Earth’s reflectance, 1999–2004, submitted to Pallé, E., P. R. Goode, P. Montañés-Rodriguez, and S. E. types and averaged in fi ve-year bands. Journal of Geophysical Research , 2005). Koonin (2004), Changes in the Earth’s reflectance Eos, Vol. 87, No. 4, 24 January 2006 over the past two decades, Science, 304, 1299–1301, cant reduction in global radiation with discussion Author Information doi:10.1126/science.1094070. of its probable causes and possible agricultural Pallé, E., P. Montañés-Rodriguez, P. R. Goode, S. E. Koonin, consequences,Agric. For. Meteorol., 107, 255. Enric Pallé, Big Bear Solar Observatory, Big Bear M.Wild, and S. Casadio (2005),A multi-data compar- Wielicki, B.A., et al. (2002), Evidence for large City, Calif.; E-mail: [email protected]; Philip R. Goode, ison of shortwave climate forcing changes, Geophys. decadal variability in the tropical mean radiative Big Bear Solar Observatory and W. K. Kellogg Radia- Res. Lett., 32, L21702, doi:10.1029/2005GL023847. energy budget, Science, 295, 841. tion Laboratory, California Institute of Technology Ramanathan,V. , R. D. Cess, E. F. Harrison, P. Minnis, Wielicki, B.A.,T. Wong, N. Loeb, P. Minnins, K. Priestley, (Caltech), Pasadena; Pilar Montañés-Rodriguez, Big B. R. Barkstrom, E.Ahmad, and D. Hartmann (1989), and R. Kandel (2005), Changes in Earth’s albedo Bear Solar Observatory; and Steven. E. Koonin,W. K. Cloud-radiative forcing and climate: Results from measured by satellite, Science, 308, 825. Kellogg Radiation Laboratory, Caltech the Earth Radiation Budget Experiment, Science, Wild, M., H. Gilgen,A. Roesch,A. Ohmura, C. Long, 243, 57. and E. G. Dutton (2005), From dimming to bright- Stanhill, G., and S. Cohen (2001), Global dimming:A ening:Trends in solar radiation inferred from sur- review of the evidence for a widespread and signifi- face observations, Science, 308, 847.

was delayed and then put on indefi nite hold news following the loss of Columbia in 2003. The satellite would provide a continuous synoptic view of the Earth and facilitate cli- mate science by measuring energy refl ected NASA Terminates Two Earth Observation Missions (albedo) over an entire hemisphere. No other PAGE 38 satellites can currently obtain these measure- important for weather and climate prediction, ments, which are needed to determine the Citing a lack of available funds, NASA has according to mission principal investigator effect of albedo on climate, according to ended support for two Earth observation mis- Dara Entekhabi, of the Massachusetts Institute DSCOVR principal investigator Francisco Vale- sions: Hydros, which never left the formulation of Technology, Cambridge. ro, of the University of California San Diego, phase, and the Deep Space Climate Observa- NASA allowed Hydros to enter the formula- Scripps Institution of Oceanography. In addi- tory (DSCOVR), which was already built and tion phase in 2003, and mission science and tion, DSCOVR would be able to monitor solar waiting for launch. engineering teams were working on algorithms activity and act as an early warning system for Richard Anthes, president of the University and instrument and spacecraft specifi cations solar storms that could damage communica- Corporation for Atmospheric Research, Boulder, until they received notice from NASA in De- tions systems on Earth, he said. Colo., said that these two new cancellations are cember 2005 that the agency would no longer NASA has already spent $100-120 million on “further evidence that…the nation’s Earth-ob- fund the project because it only had funds for DSCOVR, and Valero estimated that the launch servation satellite programs are at risk.” the two primary missions, the Orbiting Carbon and operations, including adapting it to a new Anthes co-chairs a committee of the Na- Observatory and Aquarius. NASA had spent launch vehicle, would cost another $60–120 tional Research Council (NRC) of the U.S. Na- about $5 million total on Hydros, according to million. However, he noted that the costs could tional Academies that is conducting a decadal Entekhabi. be shared between NASA and the U.S. Nation- survey of Earth science and applications from Anthes said the termination of Hydros “means al Oceanic and Atmospheric Administration, space, which should be completed by the end that [ESSP] is indefinitely delayed at best and, at which has expressed interest in the satellite’s of 2006. In a September 2005 interim report, worst, it is dead.”The 2005 NRC report had high- solar monitoring capabilities for use in disas- the committee concluded that the U.S. system lighted ESSP as a critical need for the nation’s ter prevention and has commissioned a study of Earth observation satellites was “at risk of Earth observation programs. of the agency’s possible participation. collapse,” in part because of the cancella- Jack Kaye, director of the research and Kaye said that in the current budget environ- tions or delays of six other missions (see Eos, analysis program in NASA’s Earth-Sun System ment that has left NASA with limited funds for (86)43, 2005). Division, said that NASA plans to have another Earth science, it did not make sense for NASA Hydros was selected as an alternate mission round of ESSP competition,although no date to continue supporting DSCOVR in the indeter- in 2002 in the third round of competition for has been announced. Hydros investigators can minate state it was left in following the loss of the Earth System Science Pathfi nder (ESSP) re-propose the mission in the next round of its launch vehicle. program—which funds small- and medium- ESSP competition, Kaye said. “Nobody likes to start something and sized Earth science missions—and had a ten- not be able to finish it,” he said.“But at some tative launch date of September 2010.The mis- A Satellite Grounded point, you have to manage with the budget sion would have used low frequency mapping you have and make priority-based decisions.” radar and radiometer instruments to measure DSCOVR has a longer history: In March 1998, However,Valero said that NASA has to de- soil moisture for use in weather and climate then U.S.Vice President Al Gore proposed a mis- cide to choose projects that are not cancelled prediction models, in drought mitigation sion named Triana to provide real-time images before they fly because it is a waste of money efforts, and for better understanding of the of the Earth from the Lagrange-1 (L-1) point, and the time and effort of the global hydrologic, energy, and carbon cycles. where the gravitational forces of the Earth and scientists involved. Hydros, unlike other current and planned Sun are equal. Panned by some Republicans as satellites that use high frequency ranges, ‘Goresat’, the mission was reborn as DSCOVR fol- —SARAH ZIELINSKI, Staff Writer would be able to make these measurements lowing a 2000 NRC report that noted its scientifi c under vegetation canopies and would be able importance.The satellite was originally sched- to determine if the moisture is frozen, which is uled for launch via the space shuttle in 2001 but