Recent Advances in Measuring Cloud Albedo with SatellitesA. FROSSARD, A. GOMEZ, J. DWYER, P. SHAW, R. SCHWARTZScripps Institution of Oceanography, La Jolla, CaliforniaABSTRACTRecent advances in satellite cloud albedo measurements, specifically data-assimilation algorithms, cloud profilers, radiometers, and limitations of each arereviewed. There has been a shift from single-instrument radiance measurementsto multi-platform instruments able to probe the full vertical profile of clouds. Thisparadigm aims to measure a wide range of cloud properties which control cloudalbedo. These new developments are necessary to reduce remaining uncertaintyin cloud albedo and constrain global climate models. This review found that newdevelopments, such as multi-instrument techniques, are necessary to verify results,and new instr umentation and methods enhance the accuracy of the results.1. IntroductionPlanetary albedo (α), the fraction of solarradiation incident on Earth that is immediatelyreflected to space, can be altered with changesin Earth’s cloud properties, aerosol content,and surface coverage (Wielicki et al. 2005).After accounting for reflection, the amount ofincoming solar intensity (S0) available for theEarth’s radiation budget is S0(1 − α). Equationsderived from first principles provide an approx-imate spatially averaged planetary albedo ofα =0.30 ± 0.005 (Seinfeld and Pandis 2006;Curry and Webster 1999; Wallace and Hobbs2006), of which 50-60% is contributed by clouds.By neglecting other climate feedbacks, theseequations show changes of 1% in albedo canalter Earth’s surface temperature by 1 K. Fur-thermore, a 1% change in albedo is equivalentto the effect of doubling the atmospheric con-centration of CO2(Solomon et al. 2007). Suchexamples demonstrate the importance of albedoon Earth’s climate and the importance of havingaccurate measurements (Smil 2008).A conventional planetary albedo (αp) hasbeen defined with respect to fractional cloudcover (Fcloud) in order to divide cloudy andclear sky influences of albedo(Acloudand Aclear,respectively) (Cess 1976):αp= (1 − Fcloud)Aclear+ FcloudAcloud(1)Any planetary albedo definition is incompletewithout cloud contribution. Not all clouds con-tribute equally to albedo; most clouds reflectincoming solar radiation, but some actually en-hance the greenhouse effect by trapping theoutgoing terrestrial radiation. Large low lyingclouds, such as stratocumulous, are the maincontributors to cloud albedo. High altitude, thin,icy cirrostratus clouds are responsible for trap-ping terrestrial infrared (IR) radiation (Halusa2008).The 1960 Television Infrared ObservationSatellite (TRIOGS-1) was the first successfulmeteorological satellite and lead to the long1running Nimbus program. Numerous satelliteshave made cloud optical and physical propertymeasurements since NASA’s 1964 launch of thefirst Nimbus satellite (Grayzeck 2003). Remotesensing data is not exclusive to satellites andhas, to some extent, been around for decades.Despite the relatively new availability of highquality satellite data, large uncertainties in themeasurements make validation of global cli-mate models problematic. Parameterization ofmicron-sized particles and drops onto planetaryscales makes clouds and aerosols the source oflargest uncertainty in predicted climate change(Solomon et al. 2007). A series of advancedsatellites are now measuring properties ofthese two critical atmospheric constituents toconstrain models. These newer satellites em-ploy more sophisticated technologies, includingLIDAR, RADAR, and highly sensitive radiome-ters. Satellite measurements provide global,consistent, and reliable observations that areunparalleled by ground-based instr uments. Themost recent and important satellite advancesin the measurements of cloud albedo are re-viewed, covering algorithms, cloud profilers,radiometers, and their limitations.2. AlgorithmsThe advent of multi-platform observationsystems requires retrieval algorithms that cantranslate raw data from a variety of sources intousable properties of clouds. Several satellitescarry more than one sensor per platform andother sensors are located on multiple platforms.Radiative budgets are generally calculated asfluxes and changes in temperatures, whereassatellite measurements are measured in radi-ances. Conversion algorithms are then neededto convert measurements into compatible formsfor input into climate models. The MultilayeredCloud Retrieval System (MCRS) is a recent(2004) technique that uses data from multiplesatellites to determine the characteristics ofmultilayer clouds. The older, homogeneous,single-layer assumption introduces large errorsin cloud properties and in the resulting albedodue to differences in actual composition (Yi et al.2007). To overcome the homogeneous, single-layer cloud assumption the MCRS algorithmwas designed from the results of a two-layercloud model (Huang et al. 2004). MCRS wasdeveloped to provide a more comprehensivelook at cloud structure and is an improvementupon the original single-layer algorithm, VisibleInfrared Solar-Infrared Split Window Technique(VISST).Vertical layering of ice clouds over waterclouds presents a large impediment in deter-mining overall cloud optical depth and albedo(Minnis et al. 2007). Ice water and liquid waterphases in clouds contribute differently to theoverall cloud albedo because they have differ-ent reflectivities. The optical depth of a clouddepends on the ice water path (IWP) and liquidwater path (LWP), so it is important to accu-rately measure both. Albedo increases with anincrease in the optical depth (Curry and Web-ster 1999). Optical depth, derived from reflectedvisible and infrared radiance, also includes acombination of the radiative transfer betweencloud layers. This leads to an approximate 40%overestimation when a multilayered cloud isconsidered to be a single, homogeneous layer(Huang et al. 2004).Microwave radiance data from satellites isused to directly determine the LWP of clouds(Minnis et al. 2007), defined as the verticalintegral of the liquid water mixing ratio (Curryand Webster 1999). In previous methods, cloudIWP was found by subtracting the LWP fromthe total water path (TWP) (Minnis et al. 2007).MCRS uses a parameterization of the origi-nal adding-doubling radiative transfer method.This was a statistical method which assumeda single homogeneous layer, by combining thelower layer cloud with the surface to producebackground radiance for the retrieval
View Full Document
Unlocking...