Gang Hong

CERES Edition-2 Cloud Property Retrievals Using TRMM VIRS and Terra and Aqua MODIS Data—Part I: Algorithms

The National Aeronautics and Space Administrations Clouds and the Earths Radiant Energy System (CERES) Project was designed to improve our understanding of the relationship between clouds and solar and longwave radiation. This is achieved using satellite broad-band instruments to map the top-of-atmosphere radiation fields with coincident data from satellite narrow-band imagers employed to retrieve the properties of clouds associated with those fields. This paper documents the CERES Edition-2 cloud property retrieval system used to analyze data from the Tropical Rainfall Measuring Mission Visible and Infrared Scanner and by the MODerate-resolution Imaging Spectrometer instruments on board the Terra and Aqua satellites covering the period 1998 through 2007. Two daytime retrieval methods are explained: the Visible Infrared Shortwave-infrared Split-window Technique for snow-free surfaces and the Shortwave-infrared Infrared Near-infrared Technique for snow or ice-covered surfaces. The Shortwave-infrared Infrared Split-window Technique is used for all surfaces at night. These methods, along with the ancillary data and empirical parameterizations of cloud thickness, are used to derive cloud boundaries, phase, optical depth, effective particle size, and condensed/frozen water path at both pixel and CERES footprint levels. Additional information is presented, detailing the potential effects of satellite calibration differences, highlighting methods to compensate for spectral differences and correct for atmospheric absorption and emissivity, and discussing known errors in the code. Because a consistent set of algorithms, auxiliary input, and calibrations across platforms are used, instrument and algorithm-induced changes in the data record are minimized. This facilitates the use of the CERES data products for studying climate-scale trends.

Uncertainties Associated With the Surface Texture of Ice Particles in Satellite-Based Retrieval of Cirrus Clouds: Part II—Effect of Particle Surface Roughness on Retrieved Cloud Optical Thickness and Effective Particle Size

The simplified ray-tracing technique reported in Part I of this paper is employed to compute the single-scattering properties of hexagonal columns with maximum dimensions ranging from 2 to 3500 mum with a size-bin resolution of 2 mum at wavelengths of 0.86 and 2.13 mum. For small ice crystals, the current treatment of surface roughness may not be adequate because the applicability of the principles of geometric optics breaks down for small roughness scale. However, for ice crystals smaller than 40 mum, the aspect ratios of these particles are close to one, and the effect of surface roughness is quite small. In this paper, the diffraction is accounted for in the same way as in the case of smooth particles. It is essentially unfeasible to incorporate the effect of surface roughness into the numerical computation of the diffraction contribution. The scattering properties of individual ice crystals are then averaged over 18 particle size distributions whose effective particle radii (re) range from 5 to 90 mum. The single-scattering properties of ice clouds are strongly sensitive to surface roughness condition. Lookup tables that are built for the correlation between the bidirectional reflectances at wavelengths of 0.86 and 2.13 mum with different roughness conditions are used to retrieve ice cloud optical thickness and effective particle size over oceans. Pronounced differences are noticed for the retrieved cirrus cloud optical thickness and effective particle sizes in conjunction with different surface roughness conditions. The values of the retrieved cirrus cloud optical thickness in the case of the rough surface are generally smaller than their counterparts associated with smooth surface conditions. The effect of surface roughness on the retrieved effective particle radii is not pronounced for slight and moderate roughness conditions. However, when the surfaces of ice crystals are substantially rough, the retrieved effective radii associated with roughened particles are larger and smaller than their smooth surface counterparts forlarge (re>50 mum) and small (re<35 mum) ice crystals, respectively, whereas the effect of surface roughness on the retrieved effective radii shows a nonmonotonic feature for moderate particle sizes (35 mum<re<50 mum). In general, the dominant effect of surface roughness on cloud property retrievals is to decrease the retrieved optical thickness and to increase the retrieved effective particle size in comparison with their counterparts in the case of smooth ice particles.

Uncertainties Associated With the Surface Texture of Ice Particles in Satellite-Based Retrieval of Cirrus Clouds—Part I: Single-Scattering Properties of Ice Crystals With Surface Roughness

Surface roughness of ice crystals is a morphological parameter important to the scattering characteristics of these particles. The intent of this paper, reported in two parts (hereafter, Parts I and II), is to investigate the accuracy associated with some simplifications in calculating the single-scattering properties of roughened ice crystals and to quantify the effect of surface roughness on the retrieval of the optical and microphysical properties of ice clouds from satellite observations. In Part I, two ray-tracing schemes, a rigorous algorithm and an approximate algorithm with a simplified treatment of surface roughness, are employed to calculate the single-scattering properties of randomly oriented hexagonal ice crystals with size parameters in the geometric optics regime. With the rigorous approach, it requires substantial computational effort to accurately account for the multiple external reflections between various roughness facets and the reentries of outgoing rays into the particles in the ray-tracing computation. With the simplified ray-tracing scheme, the ray-tracing calculation for roughened particles is similar to that for smooth particles except that, in the former case, the normal of the particle surface is statistically perturbed for each reflection-refraction event. The simplified ray-tracing scheme can account for most the effects of surface roughness on particle single-scattering properties without incurring substantial demand on computational resources and, thus, provides an efficient way to compute the single-scattering properties of roughened particles. The effect of ice-crystal surface roughness on the retrieval of the optical thicknesses and effective particle sizes of cirrus clouds is reported in Part II.

Differences Between Collection 4 and 5 MODIS Ice Cloud Optical/Microphysical Products and Their Impact on Radiative Forcing Simulations

This paper reports on the comparison of two latest versions (collections 4 and 5) of ice cloud products derived from the Moderate Resolution Imaging Spectroradiometer (MODIS) measurements. The differences between the bulk optical properties of ice clouds used in collections 4 and 5 and the relevant impact on simulating the correlation of the bidirectional reflection functions at two MODIS bands centered at 0.65 (or 0.86) and 2.13 mum are investigated. The level-3 MODIS ice cloud properties (specifically, ice cloud fraction, optical thickness, and effective particle size in this paper) from the collection 4 and 5 datasets are compared for a tropical belt of 30deg S-30deg N. Furthermore, the impact of the differences between the MODIS collection 4 and 5 ice cloud products on the simulation of the radiative forcing of these clouds is investigated. Over the tropics, the averaged ice cloud fraction from collection 5 is 1.1% more than the collection 4 counterpart, the averaged optical thickness from collection 5 is 1.2 larger than the collection 4 counterpart, and the averaged effective particle radius from collection 5 is 1.8 mum smaller than the collection 4 counterpart. Moreover, the magnitude of the differences between collection 5 and 4 ice cloud properties also depends on the surface characteristics, i.e., over land or over ocean. The differences of these two datasets (collections 4 and 5) of cloud properties can have a significant impact on the simulation of the radiative forcing of ice clouds. In terms of total (longwave plus shortwave) cloud radiative forcing, the differences between the collection 5 and 4 results are distributed primarily between -60 and 20 W ldr m-2 but peak at 0 W ldr m-2.

High Cloud Properties from Three Years of MODIS Terra and Aqua Collection-4 Data over the Tropics

Abstract This study surveys the optical and microphysical properties of high (ice) clouds over the Tropics (30°S–30°N) over a 3-yr period from September 2002 through August 2005. The analyses are based on the gridded level-3 cloud products derived from the measurements acquired by the Moderate Resolution Imaging Spectroradiometer (MODIS) instruments aboard both the NASA Earth Observing System Terra and Aqua platforms. The present analysis is based on the MODIS collection-4 data products. The cloud products provide daily, weekly, and monthly mean cloud fraction, cloud optical thickness, cloud effective radius, cloud-top temperature, cloud-top pressure, and cloud effective emissivity, which is defined as the product of cloud emittance and cloud fraction. This study is focused on high-level ice clouds. The MODIS-derived high clouds are classified as cirriform and deep convective clouds using the International Satellite Cloud Climatology Project (ISCCP) classification scheme. Cirriform clouds make up more tha...

Parameterization of shortwave and longwave radiative properties of ice clouds for use in climate models.

Abstract Climate modeling and prediction require that the parameterization of the radiative effects of ice clouds be as accurate as possible. The radiative properties of ice clouds are highly sensitive to the single-scattering properties of ice particles and ice cloud microphysical properties such as particle habits and size distributions. In this study, parameterizations for shortwave (SW) and longwave (LW) radiative properties of ice clouds are developed for three existing schemes using ice cloud microphysical properties obtained from five field campaigns and broadband-averaged single-scattering properties of nonspherical ice particles as functions of the effective particle size De (defined as 1.5 times the ratio of total volume to total projected area), which include hexagonal solid columns and hollow columns, hexagonal plates, six-branch bullet rosettes, aggregates, and droxtals. A combination of the discrete ordinates radiative transfer model and a line-by-line model is used to simulate ice cloud rad...

Interannual to Diurnal Variations in Tropical and Subtropical Deep Convective Clouds and Convective Overshooting from Seven Years of AMSU-B Measurements

Abstract This study surveys interannual to diurnal variations of tropical deep convective clouds and convective overshooting using the Advanced Microwave Sounding Unit B (AMSU-B) aboard the NOAA polar orbiting satellites from 1999 to 2005. The methodology used to detect tropical deep convective clouds is based on the advantage of microwave radiances to penetrate clouds. The major concentrations of tropical deep convective clouds are found over the intertropical convergence zone (ITCZ), the South Pacific convergence zone (SPCZ), tropical Africa, the Indian Ocean, the Indonesia maritime region, and tropical and South America. The geographical distributions are consistent with previous results from infrared-based measurements, but the cloud fractions present in this study are lower. Land–ocean and Northern–Southern Hemisphere (NH–SH) contrasts are found for tropical deep convective clouds. The mean tropical deep convective clouds have a slightly decreasing trend with −0.016% decade−1 in 1999−2005 while the m...

The Sensitivity of Ice Cloud Optical and Microphysical Passive Satellite Retrievals to Cloud Geometrical Thickness

Most satellite-based ice cloud retrieval algorithms rely on precomputed lookup libraries for inferring the ice cloud optical thickness (tau) and effective particle size ( De). However, this retrieval methodology does not account for the case where cloud geometrical thickness may vary by several kilometers. In this paper, we investigate the effect of the ice cloud geometrical thickness on the retrieval of tau and De for algorithms using the Moderate Resolution Imaging Spectroradiometer infrared (IR) bands at 8.5 and 11 mum (or 12 mum) or solar bands at 0.65 and 1.64 mum (or 2.13 mum). We use a rigorous radiative transfer package to simulate the IR brightness temperatures and solar reflectances, assuming that the ice cloud top height is fixed at 12 or 15 km with a variation of cloud geometrical thickness from 0.5 to 5 km. The simulated brightness temperatures and reflectances are then used to investigate the errors of cloud tau and De inferred from the precomputed lookup tables developed with a specific geometrical thickness. It is found that the retrieval errors in tau and De increase with increasing tau for the IR and solar methods. In both cases, cloud tau and De may be underestimated and overestimated, respectively, if the effect of the cloud geometrical thickness is not taken into account. The effect of the cloud geometrical thickness on the retrieval of cloud optical and microphysical properties is much larger for the IR algorithm than for the solar-band-based algorithm. This paper demonstrates that the inclusion of the information about the cloud geometrical thickness may improve the accuracy of the retrieval of the cloud properties on the basis of the precomputed lookup libraries

Contrails and Induced Cirrus: Optics and Radiation

This paper summarizes the assessment of the current state of knowledge, areas of uncertainties, and recommendations for future efforts, regarding the optical and radiative properties of contrails and contrail cirrus, which have been reported in two detailed subject-specific white papers for the Aviation Climate Change Research Initiative undertaken by the U.S. Federal Aviation Administration. To better estimate the radiative forcing of aircraft-induced cloudiness, there is a pressing need to improve the present understanding of the optical properties of nonspherical ice crystals within contrails and contrail cirrus, and to enhance the global satellite detection and retrieval of these clouds. It is also critical to develop appropriate parameterizations of ice crystal bulk optical properties for climate models on the basis of state-of-the-art scattering simulations and available in situ measurements of ice crystal size and habit distributions within contrails and contrail cirrus. More accurate methods are n...

The characterization of deep convective cloud albedo as a calibration target using MODIS reflectances

There are over 25 years of historical satellite data available for climate analysis. The historical satellite data needs to be properly calibrated, especially in the visible, for sensors with no onboard calibration. Accurate vicarious calibration of historical satellites relies on invariant targets, such as the moon, Dome C, and deserts. Deep convective clouds (DCC) also show promise of being a stable or predictable target viewable by all satellites, since they behave as solar diffusers. However DCC have not been well characterized for calibration. Ten years of well-calibrated MODIS radiances are now available. DCC can easily be identified using IR thresholds, where the IR calibration can be traced to the onboard blackbodies. The natural variability of the DCC radiance will be analyzed geographically, seasonally, and for differences of convection initiated over land and ocean. Functionality between particle size and ozone absorption with DCC albedo will be examined theoretically. Although DCC clouds are nearly Lambertian, the angular distribution of reflectances will be sampled and compared with theoretical models. Both Aqua and Terra MODIS DCC angular models were compared for consistency. The DCC method was able to identify two calibration coefficient discontinuities in the Terra-MODIS Collection 5 10-year record and validated the calibration stability of MODIS to within 0.1% per decade. The DCC method needs to take into account the functionality of the 0.65μm DCC radiance with the 11μm brightness temperature threshold and the DCC 0.65μm radiance difference observed over the tropical western pacific and the afternoon generated DCC over land. Both of these cases cause a bias on the order of 5%. These improvements are the first steps towards successful use of DCC as an absolute calibration target.