Clouds and the Earth’s Radiant Energy System (CERES) Cloud Radiative Swath (CRS) Edition 4 Data Product
1. Introduction
The Clouds and the Earth’s Radiant Energy System (CERES) project seeks to quantify the radiation budget of the entire Earth–atmosphere system, from the top of the atmosphere (TOA), where radiation governs the total amount of energy contained within the global climate system, to within the atmosphere and at the surface, where radiant energy exchanges control fundamental climate processes and feedbacks. CERES radiometers measure reflected solar and emitted thermal infrared energy exiting the Earth–atmosphere system from four sun-synchronous satellites, namely, the Terra, Aqua, Suomi NPP, and NOAA-20 spacecraft. Together with data from coincident high-resolution spectral imagers, these measurements have established an accurate multidecadal record of TOA radiative fluxes and associated cloud and aerosol properties for Earth radiation budget and climate studies. Yet deciphering the energetics and associated dynamics of Earth’s climate requires information on radiant energy flows internal to the system, especially at the surface where radiation contributes to land and ocean heating, the hydrological cycle, snow and ice melt, and water vapor and ice-albedo feedbacks. While surface radiative fluxes can be observed locally by surface radiometric instrumentation, obtaining such information globally is only possible via computations.
To supplement the CERES satellite observations, surface broadband shortwave (SW) and longwave (LW) irradiances are estimated within CERES instrument footprints using three simple parameterization algorithms (Kratz et al. 2010, 2020). These instantaneous parameterized surface fluxes are reported with the TOA satellite observations in the CERES level 2 (L2) Single Scanner Footprint (SSF) TOA/Surface Fluxes and Clouds Edition 4 (Ed4) data product. Most notable are the all-sky Langley Parameterized SW and LW Algorithms (LPSA and LPLA, respectively), collectively referred to as Model B (Gupta 1989; Gupta et al. 1992; Gupta 2001). Additional surface flux parameterizations used in CERES SSF processing include Model A (Li and Garand 1994; Inamdar and Ramanathan 1997), which estimates fluxes in clear-sky footprints only, and Model C (Zhou and Cess 2001; Zhou et al. 2007), which was recently incorporated into the CERES Ed4 processing stream (Kratz et al. 2020) and estimates LW fluxes only. However, these parameterization algorithms have several limitations in scope and accuracy. While Model B has undergone various refinements over the years (e.g., Gupta et al. 1992, 2010), Models A and C have not seen significant improvements since their inception. Hence, there is a need to improve the CERES L2 radiative fluxes to facilitate accurate process-level radiation budget studies, climate data algorithm improvements, and applied sciences, among other applications. Explicitly calculating radiative fluxes using a radiative transfer model (RTM) is expected to yield more accurate results (e.g., Huang et al. 2020) and can provide a more sophisticated and useful suite of CERES L2 flux products.
In this paper, we describe the CERES L2 Cloud Radiative Swath (CRS) Ed4 data product and evaluate its performance against available ground-based measurements and CERES TOA observations during the period 2019/20. Originally developed by the CERES science team in the 2000s, CRS extends the standard L2 SSF data product by calculating radiative fluxes at the instantaneous footprint scale using the NASA Langley Fu–Liou (LFL) RTM (Fu and Liou 1993; Kratz and Rose 1999; Kato et al. 1999, 2005). CRS data were previously released with CERES Edition 2 (Ed2) data products (Rutan et al. 2009; Rose et al. 2013). However, CRS Ed2 data production ceased shortly thereafter owing to high computational cost and a greater need to prioritize the development of CERES level 3 (L3) gridded data products used for climate model evaluation (Doelling et al. 2013; Rutan et al. 2015; Boeke and Taylor 2016; Doelling et al. 2016; Kato et al. 2018; Loeb et al. 2018a). The CRS Ed4 framework enables the CERES science team to perform radiative closure studies to evaluate new cloud and aerosol retrievals (e.g., Loeb et al. 2018b), and to diagnose changes in the input datasets used to process the CERES L3 Synoptic 1-degree (SYN1deg) product (Rutan et al. 2015) and the Energy Balanced and Filled Surface (EBAF-Surface) global climate data record (Kato et al. 2018). CRS was initially developed for the cross-track scanning CERES Terra FM1 and Aqua FM3 instruments and has been produced for these instruments only.
Following a description of the CERES SSF Ed4 data product that serves as the basis for the CRS calculations, section 2 describes the CRS Ed4 flux algorithm including the inputs and outputs of the LFL RTM. In section 3, we evaluate the performance of the CRS surface downwelling broadband radiative fluxes against a globally distributed network of ground-based irradiance measurements. In addition, we compare its performance to that of the parameterized surface fluxes contained within the SSF Ed4A data product and its near-real-time counterpart, the Fast Longwave and Shortwave Fluxes (FLASHFlux) SSF version 4A (V4A) data product (Kratz et al. 2014). Using data from remote field campaigns, section 4 further evaluates CRS while highlighting its utility for scientific investigations. Section 5 evaluates the CRS computed TOA outgoing longwave radiation (OLR) and reflected shortwave radiation (RSW) fluxes against CERES SSF Ed4A observations. Section 6 summarizes our results and conclusions and provides an outlook on future CRS algorithm development.
2. Data and methods
a. CERES SSF data product
Largely anchored by the Terra and Aqua spacecraft, the CERES record now spans over two decades. Terra, launched in 1999, carries the CERES FM1 and FM2 instruments. Aqua, launched in 2002, carries the CERES FM3 and FM4 instruments. Terra descends across the equator around 1030 mean local time, while Aqua ascends across the equator around 1330 mean local time. Each instrument measures filtered broadband radiances in shortwave (SW; 0.3–5.0 μm), total (TOT; 0.3–200 μm), and window (8.0–12.0 μm) channels; LW radiances are inferred from the TOT minus SW channels. The observed filtered radiances are unfiltered using the latest CERES gains and spectral correction coefficients (Loeb et al. 2001, 2016). The unfiltered radiances are converted to outgoing radiative fluxes using empirical angular distribution models (ADMs) that depend on the observed scene type (Su et al. 2015a,b, 2020). The Terra FM1 and Aqua FM3, used for routine scientific observations, operate in a fixed azimuth plane or cross-track scan mode in which scan lines are perpendicular to the satellite ground path. To enhance angular sampling for the development of ADMs for radiance-to-flux conversion, the Terra FM2 and Aqua FM4 operated in a rotating azimuth plane scan mode prior to 2005 but have since operated largely in cross-track mode. The location, size, and shape of each footprint on Earth’s surface is determined by the instrument instantaneous field-of-view. At nadir viewing geometries, footprints are approximately circular with a nominal 20-km diameter. At oblique view zenith angles (VZAs), footprints increase in size and ellipticity. Narrowband radiances from coincident 1-km Moderate Resolution Imaging Spectroradiometer (MODIS) pixels are used for scene identification and cloud property retrievals (Minnis et al. 2020). Each footprint may contain up to three subfootprint areas, including a clear-sky area and up to two nonoverlapping cloud layers, including a lower cloud layer and an upper cloud layer. Each cloudy pixel is assigned a basic set of cloud properties. Meteorological information used for ADM selection, cloud retrievals, and surface flux estimation comes from the Goddard Earth Observing System version 5.4.1 (GEOS 5.4.1), a gridded 1° × 1° global reanalysis (Rienecker et al. 2008) produced by the Global Modeling and Assimilation Office (GMAO). Aerosol optical depth (AOD) comes from the MODIS MOD04/MYD04 product (Remer et al. 2005; Ignatov et al. 2005; Levy et al. 2013) supplemented by daily mean simulations from the Model for Atmospheric Transport and Chemistry (MATCH), a 1° × 1° data assimilation system (Collins et al. 2001). Cloud and aerosol statistics (mean, standard deviation) in each footprint are computed by weighting MODIS pixels by the energy distribution within each footprint using the CERES point spread function.
Separate from CERES SSF Ed4 processing, the FLASHFlux working group expedites SSF processing to provide fluxes to the applied science community within 4 days of the CERES satellite observations (Kratz et al. 2014). These near-real-time fluxes are useful for clean energy, infrastructure energy use, and agricultural applications. Like the CERES SSF, FLASHFlux SSF surface fluxes are estimated using the Model B parameterization algorithm. Atmospheric state information for cloud retrievals and ADM selection comes from GMAO’s near-real-time GEOS Forward Processing for Instrument Teams (FP-IT) reanalysis (Lucchesi 2013). AOD is represented using a monthly climatology from MATCH. Differences in CERES SSF and FLASHFlux SSF fluxes may arise from differences in clouds, aerosols, and/or meteorology.
b. CRS input data
The main inputs to the CRS LFL RTM include vertical profiles of temperature, pressure, humidity, ozone, and aerosols, surface temperature, albedo, and emissivity, and cloud properties (Fig. 1). Cloud properties, including the cloud percent coverage, thermodynamic phase (ice or liquid), optical depth, particle size, and top and base pressure, represent footprint-average values from the SSF Ed4 data product (Minnis et al. 2020). Atmospheric profiles of temperature, pressure, humidity, and ozone are specified using 6 hourly fields from the GEOS 5.4.1 reanalysis. Hourly MATCH simulations of dust (four size categories), sulfate, sea salt, hydrophilic black and organic carbon, hydrophobic black and organic carbon, and volcanic aerosols are used to specify the vertical distribution of aerosols. These aerosol constituents are combined into seven categories in the Optical Properties of Aerosols and Clouds (OPAC) database (Tegen and Lacis 1996; Hess et al. 1998), which provides the corresponding values of spectral single-scatter albedo and asymmetry parameter. The total AOD comes from MODIS MOD04/MYD04 clear-sky retrievals (Levy et al. 2013), when available, and otherwise comes from MATCH; in either case, MATCH determines the vertical distribution of aerosols.
Illustration of the CERES CRS Ed4 flux algorithm inputs and outputs.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Surface skin temperature comes from clear-sky pixel CERES MODIS Ed4 retrievals (Minnis et al. 2020). If a footprint is fully obscured by clouds, the surface skin temperature is instead specified from GEOS 5.4.1. Rutan et al. (2009) provide a comprehensive overview of the CRS treatment of surface albedo. Clear-sky albedo of land and snow/ice surfaces is retrieved using a scene-dependent lookup table (LUT) that uses the CERES TOA RSW flux and ancillary meteorological parameters as input (Rutan et al. 2006, 2009). Ice-free ocean surface albedo is estimated using a LUT developed using the Coupled Ocean Atmosphere Radiation Transfer Model of Jin et al. (2004) that includes dependence on surface wind speed. The surface albedo in cloudy footprints is estimated using a monthly surface albedo history map derived from SSF clear-sky MODIS pixels. Surface albedo spectral shape (Rutan et al. 2009) is determined for each LFL RTM band using the MODIS MCD43 BRDF/Albedo data product (Schaaf et al. 2002). Surface spectral emissivity is determined based upon the International Geosphere-Biosphere Programme (IGBP) surface type classification.
c. Langley Fu–Liou radiative transfer model calculations
The LFL RTM is a correlated-k radiative transfer code that accounts for scattering and absorption by cloud and aerosol particles, Rayleigh scattering, and gaseous absorption by H2O, CO2, O3, O2, CH4, NO2, and CFCs (Fu and Liou 1993; Kratz and Rose 1999; Kato et al. 1999, 2005). In a comparison of contemporary radiative transfer codes, Oreopoulos et al. (2012) found that the LFL RTM outperforms many similar RTMs. SW calculations use a 4-stream approximation to the radiative transfer equation with 18 spectral bands covering 0.175–4.0 μm (Table 1). LW calculations, which include scattering effects, use a 2-stream approximation with 12 spectral bands covering 0–2200 cm−1. The calculations are executed at high vertical resolution with up to 36 pressure levels per footprint. Water cloud optical properties are based on Mie calculations (Hu and Stamnes 1993). Ice cloud optical properties are represented using rough hexagonal columns consistent with the CERES MODIS Ed4 retrievals. Additional levels are inserted into the model at the cloud-top and cloud-base pressure. When a footprint contains multiple subfootprint areas, the RTM is run for each sky condition assuming 100% coverage of the footprint. The footprint flux is then calculated as the area-weighted mean flux. On a given day, CRS production for a single instrument entails performing multiple LFL RTM calculations within ∼2.3 × 106 CERES footprints.
Spectral bands for which narrowband all-sky radiative fluxes are provided in the CERES CRS Ed4 product, including downwelling (↓) fluxes at the surface and upwelling (↑) fluxes at TOA. SW bands 4–6 cover photosynthetically active radiation (PAR). LW bands 5–7 cover the IR atmospheric window.
Previous CRS editions used an objective constrainment algorithm to “tune” the OLR and RSW fluxes from an initial run of the LFL RTM in order to improve agreement with CERES observations. This tuning process was accomplished by altering various LFL RTM inputs within their estimated range of uncertainty. Despite improving agreement with CERES TOA observations, the CRS tuned fluxes rarely matched the CERES TOA fluxes. Rose et al. (2013) also found that TOA tuning failed to improve agreement between CRS-simulated and ground-measured surface fluxes. This is perhaps unsurprising in the LW since OLR and surface LW↓ fluxes are often decoupled. As our principal motivation is to improve the accuracy of the CERES L2 surface radiative fluxes, the CRS Ed4A calculations discussed herein consist of a single “untuned” run of the LFL RTM. Eliminating tuning drastically reduces the amount of time required for CRS processing. Hence, the only way CERES observations enter CRS Ed4 processing is when RSW is used to estimate clear-sky land and snow/ice surface albedo, as described above.
d. CRS output data
The CRS product contains vertical profiles of upwelling (↑) and downwelling (↓) broadband SW (0.175–4.0 μm) and LW (0–2200 cm−1) radiative fluxes reported at 6 standard atmospheric pressure levels, including the surface, 850, 500, 200, 70 hPa, and TOA (0.1 hPa). To enable users to estimate cloud and aerosol direct radiative effects, the flux profiles are computed for all-sky (with clouds and aerosols), clear-sky (without clouds but with aerosols), pristine sky (no clouds, no aerosols), and all-sky no aerosol (with clouds but not aerosols) conditions. CRS also contains spectrally resolved (i.e., narrowband) all-sky downwelling fluxes at the surface and TOA upwelling fluxes (Table 1). The first 4 out of 18 SW bands used in the RTM calculations are combined into a single band for output (band 1). Spectral SW outputs include ultraviolet (UV; Su et al. 2005; bands 1–2) and photosynthetically active radiation (PAR; Su et al. 2007; bands 4–6) fluxes. Direct and diffuse components of the surface broadband SW↓ flux are also provided for each sky condition mentioned above.
3. Surface radiative flux validation and comparison to SSF parameterizations
Figure 2 shows the network of ground-based irradiance measurements used to evaluate the performance of the CRS-simulatedsurface downwelling broadband irradiances. Representing a collection of coastal, desert, island, continental, and polar surface types, this network includes 40 sites (Table S1 in the online supplemental material) maintained by the international Baseline Surface Radiation Network (BSRN; Ohmura et al. 1998), NOAA’s Global Monitoring Division (GMD) Surface Radiation Network (SURFRAD; Augustine et al. 2000), and the U.S. Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) program (Stokes and Schwartz 1994). All surface irradiance measurements have a temporal resolution of 1 min. Measurements from ocean buoys are not used due to coarse hourly resolution. To quantify whether CRS fluxes are an improvement over their parameterized counterparts, we also compare CRS validation results with those for the CERES SSF Ed4A and FLASHFlux SSF V4A data products. We primarily focus on comparisons with the Model B LPSA (Gupta 2001) and LPLA (Gupta 1989; Gupta et al. 1992), since it is the most accurate and only algorithm that provides all-sky SW↓ and LW↓ fluxes [for completeness, comparisons with clear-sky Model A (Li and Garand 1994; Inamdar and Ramanathan 1997) and all-sky LW Model C (Zhou and Cess 2001; Zhou et al. 2007) are included in the supplemental materials]. The LPSA estimates the surface broadband SW↓ flux by accounting for transmission of the incoming solar beam through clear and cloudy atmospheric layers (Gupta 2001). The LPLA estimates the surface broadband LW↓ flux in terms of an effective atmospheric emission temperature, cloud amount and base height, and subcloud column-integrated water vapor (Gupta 1989; Gupta et al. 1992). Since footprint-scale (∼20–50 km) surface heterogeneity is often inconsistent with the scene observed by downward-viewing radiometers, no effort was made to validate the CRS upwelling and net surface radiative fluxes.
Map showing the location and surface-type classification of the ground-based radiation measurement sites used for surface flux validation.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
If the total SW measurement is unavailable, we resort to the global (i.e., hemispheric) SW observation from an unshaded PSP. Global measurements are given secondary priority as they are susceptible to sensor cosine response and solar heating errors. To reduce noise in the comparisons, we average the SW measurements over a ±15 min window centered at the time of the CERES footprint. We account for the changing solar zenith angle during the averaging period by scaling the footprint instantaneous SW flux by the ratio of the mean observed cosine solar zenith angle,
(top),(middle) Time series of surface irradiance measurements and collocated CERES Aqua FM3 CRS computed footprint-level surface fluxes at Payerne, Switzerland, during June 2019. (bottom) The ground-measured direct and diffuse components of the total SW↓ flux [Eq. (1)].
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Figure 4 compares the performance of the all-sky broadband SW↓ surface fluxes from CRS and the LPSA in both SSF products at the entire network of sites shown in Fig. 2. The instantaneous flux bias or difference Δ is computed as the satellite footprint minus ground-measured flux. CRS exhibits a moderate negative mean bias, whereas SSF Ed4A shows a smaller negative mean bias and FLASHFlux SSF V4A shows a relatively large positive mean bias. However, the CRS flux difference histogram exhibits a strong central peak with an SW↓ RMSΔ that is 47.7 W m−2 (33.6%) lower than CERES SSF Ed4A and 59.2 W m−2 (38.6%) lower than FLASHFlux SSF V4A. CRS SW↓ fluxes also correlate more strongly with the measured fluxes than their parameterized counterparts do. These statistics are further decomposed by surface type and cloud conditions in Table 2. CRS produces an improved SW↓ RMSΔ and linear correlation coefficient (r) over every surface type, regardless of cloud conditions. Strong improvements are found at the polar sites, where all-sky RMSΔ and r are enhanced by over 60% relative to the SSF products. Under overcast skies, drastic improvements in the mean SW↓ bias are found over the continental and coastal sites. CRS also shows the smallest overcast-sky median bias over every surface type. Although CRS is less biased than the LPSA at the polar and island sites, large biases remain owing to errors in simulated cloud transmission. At the polar sites, a arge negative mean bias under overcast skies is linked to a positive bias in CERES MODIS Ed4 cloud optical depths over polar snow/ice. Fortunately, the CERES MODIS Edition 5 (Ed5) cloud retrievals are expected to produce more realistic polar cloud optical depths, which will increase the amount of SW↓ energy reaching polar snow/ice in the future CRS Ed5 product. Major improvements are also found in comparisons to the clear-sky SW Model A (supplementary materials). In these clear-sky comparisons, CRS has an SW↓ mean bias near 0 W m−2 (Table S2), down almost 20 W m−2 in absolute value relative to Model A. RMSΔ also drops by over 67% globally, and improvements are seen over each individual surface type.
(left) Comparison of the CERES Aqua FM3 CRS daytime all-sky surface SW↓ fluxes to measurements from a globally distributed network of surface observation sites (Fig. 2) during 2019/20. Results are also presented for (center) the SW Model B (LPSA) surface flux parameterization algorithm from the CERES SSF Edition 4A and (right) FLASHFlux SSF version 4A. (top) Joint histograms indicating the number of values within each bin. (bottom) Histograms of the instantaneous flux difference Δ. N denotes the total number of samples.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Breakdown of the Aqua FM3 CRS and LPSA (Model B) surface broadband SW↓ flux validation statistics according to surface type and cloud conditions. In cells with multiple entries, the values correspond to CERES CRS Ed4A, CERES SSF Ed4A, and FLASHFlux SSF V4A, respectively. The dataset with the best statistic is highlighted in boldface font. The mean flux, mean and median flux bias, and RMS flux difference are expressed in units of W m−2. The correlation (r) is the Pearson linear correlation coefficient. The global category combines data from all sites. Overcast-sky comparisons include footprints with a cloud fraction > 95% in all products. Clear-sky comparisons include footprints with a cloud fraction < 5% in all products.
Figures 5 and 6 compare the CRS and LPLA all-sky broadband LW↓ surface flux performance at the network of ground-based sites in Fig. 2 during day and night, respectively. During daytime CRS shows the smallest mean bias and an LW↓ RMSΔ 5.4 W m−2 (19.5%) lower than CERES SSF Ed4A and 5.1 W m−2 (18.5%) lower than FLASHFlux SSF V4A (Fig. 5). At night CRS shows a negative mean bias that while larger in magnitude is comparable to the SSF products (Fig. 6). However, the nighttime LW↓ RMSΔ is 2.8 W m−2 (9.4%) lower than CERES SSF Ed4A and 3.2 W m−2 (10.6%) lower than FLASHFlux SSF V4A (Fig. 6). CRS also shows slightly higher correlations with measured LW↓ fluxes. As shown in Table 3, improvements in the LW↓ RMSΔ and r are found over every surface type, independent of cloud conditions, the only exception being clear-sky polar fluxes at night (Table 4). Negative nighttime LW↓ biases are likely in part linked to poor cloud optical thickness information, which is difficult to retrieve with only infrared channels and translates to overestimates of cloud-base height (Minnis et al. 2020; Yost et al. 2020).
(left) Comparison of CERES Aqua FM3 CRS daytime all-sky surface LW↓ fluxes to measurements from a globally distributed network of sites (Fig. 2) during 2019/20. For comparison, results are also presented for (center) the LPLA (LW Model B) surface flux parameterization algorithm from the CERES SSF Ed4A and (right) FLASHFlux SSF V4A. (top) Joint histograms indicating the number of values within each bin. (bottom) Histograms of the instantaneous flux difference Δ. N denotes the total number of samples.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
As in Fig. 5, but for nighttime surface LW↓ fluxes.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Whether CRS is less biased than the LPLA depends on the surface type, cloud conditions, and whether it is day or night. For instance, CRS exhibits large reductions in overcast-sky mean LW↓ bias during daylight hours at the desert and island locations, but it has the largest clear-sky LW↓ bias at the polar sites. This systematic bias in thermal emission at the polar sites, evident at the lowest flux values in Figs. 5 and 6, likely results from a cold-bias in the GEOS 5.4.1 near surface temperature profile, which likely fails to resolve the sharp surface-based thermal inversions that prevail over polar snow/ice under clear skies. In an attempt to reduce this bias, the LPLA adjusts its effective atmospheric emission parameter by prescribing a lower-tropospheric lapse rate of −10 K (100 hPa)−1 (Gupta et al. 2010; Kratz et al. 2020). A similar bias correction will be developed and implemented in the future CRS Ed5 product. CRS also shows the largest daytime clear-sky LW↓ bias of each product at desert sites, which may be related to errors in dust aerosol absorption/emission (e.g., Dubovik et al. 2002). Nonetheless, it has the lowest daytime clear-sky LW↓ bias over all other surface types. Similar results are found for clear-sky LW Model A and all-sky LW Model C (supplemental materials). CRS has a lower daytime bias than clear-sky LW Model A everywhere except the desert sites. Compared to clear-sky Model A and all-sky Model C, LW RMSΔ and correlations are enhanced over virtually every surface type.
4. Applications to process studies of the surface radiation budget
The CRS high-spatial-resolution instantaneous fluxes contain a wealth of information for process-level radiation budget studies. Example applications include estimating cloud and aerosol direct radiative effects, supporting field campaign research, analyzing synoptic weather and climate events, and evaluating UV and PAR surface fluxes (Su et al. 2007) relevant to ecosystems, agriculture, marine phytoplankton, and other forms of biology. The CRS atmospheric fluxes could also supplement measurements collected during airborne field campaigns. In this section, aided by measurements from two remote polar field campaigns, we further evaluate CRS while highlighting its utility for investigations of the surface radiation budget.
a. MOSAiC
The Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition observed the coupled Arctic climate system from the Polarstern icebreaker and ambient sea ice from September 2019 to October 2020 (Shupe et al. 2022). Measurements included broadband surface radiation budget components that can be used with CRS to estimate the impacts of clouds and aerosols on the surface energy budget. Figures 7 and 8 show CRS footprint-level surface fluxes collocated with the MOSAiC observatory during June 2020, as the Polarstern vessel drifted from north to south. CRS closely tracks the observed variations in surface radiative fluxes, exhibiting small LW↓ and SW↓ monthly mean bias and RMSΔ values. The CRS surface downwelling cloud radiative effect (CRE), computed as the all-sky minus clear-sky downwelling flux at the surface, has monthly means of 58.9 W m−2 in the LW and −155.5 W m−2 in the SW. Excessive polar cloud optical depths however may cause overestimation of the SW↓ CRE. Along with the detailed ship measurements, CRS can be used to assess how Arctic CREs vary as a function of cloud properties and large-scale meteorology.
Comparison of the CERES Terra FM1 and Aqua FM3 CRS surface LW↓ fluxes with MOSAiC ship-based measurements over the Arctic Ocean during June 2020. (top left) A map of CRS LW↓ surface fluxes collocated with the Polarstern ship track. (bottom) The ship-based PIR measurements along with the same CRS footprint flux values shown in the top-left panel. (top right) A scatterplot of CRS vs MOSAiC LW↓ fluxes along with the mean bias, the observed flux averaged over all collocated measurements, the RMSΔ, and the correlation (r) of the computed and observed fluxes.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
As in Fig. 7, but for surface SW↓ fluxes. Measurements are from an unshaded PSP.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
b. AWARE
The ARM West Antarctic Radiation Experiment (AWARE) campaign (Lubin et al. 2020) observed the surface energy balance on the West Antarctic Ice Sheet (WAIS) during one of the most prominent Antarctic surface melt events on record (Nicolas et al. 2017; Scott et al. 2019). On 10 January 2016, as an atmospheric river made landfall, a large spike in surface LW↓ radiation contributed to rapid and widespread melting of the Ross Ice Shelf and adjacent portions of the WAIS. As shown in Fig. 9, this impulse of LW↓ energy is well captured by the CRS product, which has a small mean LW↓ bias of −3 W m−2 during the period shown. LW↓ fluxes are systematically underestimated under clear-skies but show better cloudy-sky performance. On the other hand, while instantaneous clear-sky SW↓ flux errors are small (e.g., on 4 and 5 January), a mean SW↓ bias of −42 W m−2 is consistent with excessive polar cloud optical depths (e.g., 7 and 12 January). Wilson et al. (2018) observed a shift toward more frequent and optically thicker low-level liquid-bearing clouds during the melt event. As shown in Fig. 10, this shift manifested as enhanced LW↓ CRE at the surface, representing a radiative warming effect that aided surface melting.
Comparison of the CERES Terra FM1 and Aqua FM3 CRS surface downwelling radiative fluxes with AWARE measurements at WAIS Divide, Antarctica (79.467°S, 112.085°W) during January 2016. CRS fluxes are colored by the CERES MODIS Ed4 footprint cloud fraction. Orange shading highlights the melt event.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Normalized histograms of surface LW↓ CRE over WAIS Divide before/after and during the January 2016 surface melt event (Nicolas et al. 2017). These calculations include footprints from Aqua and Terra and do not include any of the clear-sky footprints shown in Fig. 9.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
5. Evaluation of CRS TOA OLR and RSW fluxes against CERES observations
We now turn to the TOA and compare the CRS computed OLR and RSW fluxes to CERES TOA observations. Figures 11 and 12 present time series of global-scale validation statistics for 2019/20, along with a representative daily map and histogram of the instantaneous ΔOLR for daytime and nighttime, respectively. CRS OLR is negatively biased, indicating insufficient LW↑ emission to space, but remains within 1% (2–3 W m−2) of the observed OLR (Figs. 11c and 12c). Much of the negative OLR bias is associated with clear-sky footprints, which suggests that the GEOS 5.4.1 atmosphere may be too cold and/or moist. However, the overall negative bias is partially compensated by a positive bias associated with high-level ice clouds (Fig. S3). Figure 13a shows the mean ΔOLR as a function of cloud effective temperature and cloud optical depth, which confirms that such clouds emit or transmit too much LW energy to space. Through comparisons with satellite active sensor retrievals, Yost et al. (2020) find that the CERES MODIS Ed4 algorithm underestimates deep convective and thin ice cloud-top heights by 700 m to 2.2 km, consistent with a warm bias in cloud radiating temperature and a positive bias in OLR. The planetary all-sky RMSΔ remains fairly stable between approximately 8 and 9 W m−2, and the computed and observed OLR fluxes are highly correlated. A slight discontinuity in clear-sky OLR bias and RMSΔ is related to microwave humidity sounder instrument loss and subsequent lack of data assimilation in GEOS 5.4.1 in November 2019.
CERES Aqua FM3 CRS daytime OLR validation statistics relative to SSF Ed4A observations. (a) A map of the instantaneous CRS computed OLR minus the SSF Ed4A observed OLR on 1 Jan 2019. (b) A histogram of the instantaneous flux bias ΔOLR. Also shown are time series of daily validation statistics for the period 1 Jan 2019–31 Dec 2020, including (c) the planetary-mean absolute bias (red) and relative bias (blue), and (d) the RMSΔ (cyan) and Pearson linear correlation coefficient
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
As in Fig. 11, but for the nighttime TOA OLR.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Monthly mean TOA OLR (day and night) and RSW bias (CRS minus SSF Ed4A) as a function of cloud effective temperature and cloud optical depth for the Aqua FM3 instrument during January 2019.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
Similar comparisons of the computed and observed TOA RSW fluxes are shown in Figs. 13b and 14. The all-sky CRS RSW exhibits a daytime global mean bias of ∼2%–5% (6–11 W m−2; Fig. 14c). Cloud reflection errors, which are most pronounced for high-level ice clouds, are the primary driver of this bias (Fig. 13b, Fig. S3). Ham et al. (2022) attribute excessive cloud SW reflection to a small-bias in the CERES MODIS Ed4 ice particle size retrievals. The ice crystal habit model used in the calculations may also play a role (Loeb et al. 2018b) and will be explored further prior to CRS Ed5. Cloud reflection errors may also arise from partly cloudy pixels treated as fully cloudy (Ham et al. 2019), or from ignoring the horizontal variability of cloud optical thickness. Clear-sky footprint RSW fluxes exhibit better performance, with a stable planetary mean bias near 0 W m−2 (1%). The global RMSΔ stays in the vicinity of 30 W m−2 and peaks during April, which may involve surface albedo errors over continents in the Northern Hemisphere. The correlation among computed and observed RSW fluxes nonetheless remains high.
As in Fig. 11, but for the TOA RSW.
Citation: Journal of Atmospheric and Oceanic Technology 39, 11; 10.1175/JTECH-D-22-0021.1
6. Summary and conclusions
To provide more accurate and advanced CERES L2 flux products, after over a decade since production ceased for CRS Ed2 (Rutan et al. 2009; Rose et al. 2013) efforts have been underway to modernize the CRS product. This paper described and validated the CRS Ed4 flux algorithm using available ground-based measurements and CERES observations, and also compared its performance with that of the parameterized surface-only flux algorithms (Kratz et al. 2020). Using the LFL RTM, CRS builds and improves upon the standard CERES SSF product by calculating instantaneous broadband and spectrally resolved radiative fluxes within CERES footprints (∼20–50 km) from the cross-track scanning Aqua FM3 and Terra FM1 instruments. The radiative transfer calculations use cloud properties from the CERES MODIS Ed4 retrieval algorithm, meteorological assimilation data from GEOS 5.4.1, aerosol information from MODIS and MATCH, and ancillary sources. The CRS product contains vertical profiles of broadband SW↓↑ and LW↓↑ radiative fluxes at 6 atmospheric pressure levels. To facilitate estimation of cloud and aerosol direct radiative effects, the flux profiles are provided for all-sky, clear-sky, pristine-sky, and all-sky no-aerosol conditions. The CRS product also contains spectrally resolved all-sky fluxes in 14 SW bands and 12 LW bands at the surface (↓) and TOA (↑), as well as direct and diffuse components of the broadband SW↓ flux at the surface. All fluxes are reported at the native CERES footprint resolution in ungridded form, with no temporal interpolation or spatial averaging.
Validation of the CRS flux algorithm for a 2-yr period (2019/20) relative to measurements from a network of 40 ground-based sites demonstrates that CRS surface fluxes represent a considerable improvement over the surface-only flux algorithms (Models A, B, C) used in CERES SSF Ed4 processing. Relative to the LPSA (SW Model B), CRS all-sky SW↓ fluxes show a ∼30%–40% reduction in RMSΔ, enhanced correlation with observations, and the smallest bias over most surface types. SW↓ RMSΔ and correlation enhancements are found under overcast and clear-sky conditions over all five surface types. Even stronger improvements are found relative to the clear-sky SW Model A (67% drop in RMSΔ, elimination of a large systematic bias). Improvements in surface LW↓ validation statistics tend to be most pronounced during daytime. Relative to the LPLA (LW Model B), CRS all-sky LW↓ fluxes show the smallest mean bias, a ∼20% reduction in RMSΔ during daytime (versus ∼10% at night), and slight increases in correlation. LW↓ RMSΔ and correlation enhancements occur under overcast and clear-sky conditions for every surface type, except at polar sites during nighttime under clear skies. Similar improvements are found relative to LW Models A and C. Yet several outstanding flux biases for specific surface types and sky conditions point to a need to further refine the CRS algorithm. Despite being biased lower than the LPSA over polar sites and islands, CRS SW↓ fluxes still show large biases linked to errors in cloud transmission. Similarly, CRS clear-sky LW↓ fluxes are systematically biased low at polar sites, for which a correction will be developed and implemented prior to Ed5. Additional comparisons at two remote polar locations demonstrated the utility of CRS fluxes for process-scale radiation budget studies, while also highlighting some of the CRS Ed4 strengths and weaknesses. Global comparisons of CRS computed TOA outgoing radiative fluxes against the CERES-derived SSF Ed4 fluxes indicate reasonable and fairly stable performance. The daily global mean OLR bias stays within 1% of CERES observations, whereas the global daily RSW bias ranges between ∼2% and 4.5% due largely to excessive reflection by clouds.
This study represents a positive step in a long history of effort to produce high quality CERES L2 surface radiative fluxes (Gupta 1989; Gupta et al. 1992; Gupta 2001; Rutan et al. 2009; Rose et al. 2013; Kratz et al. 2020). Given their overall superiority, the CRS fluxes will replace the parameterized surface fluxes on the SSF for CERES Ed5. The CRS code has additional benefits for CERES radiation algorithm development. For instance, it enables radiative closure studies to be performed to assist in future development of CERES MODIS cloud retrieval algorithms. As the Terra and Aqua satellites drift and are eventually decommissioned, development of a CERES NOAA-20 FM6 CRS product based upon Visible Infrared Imaging Radiometer Suite (VIIRS) cloud, aerosol, and surface properties is planned to ensure continuity of high-quality RTM-based L2 fluxes.
Acknowledgments.
This research was supported by the NASA CERES project. We thank Lazaros Oreopoulos and two anonymous reviewers for providing constructive comments on our initial manuscript.
Data availability statement.
The CERES CRS Ed4 data will be made available for a 5-yr period (2017–21) on the CERES ordering tool (https://ceres.larc.nasa.gov/data/). The CERES SSF Ed4A and FLASHFlux SSF V4A data are available online (https://ceres.larc.nasa.gov/products-info.php?product5SSF-Level2). MOSAiC data are available at https://doi.org/10.18739/A20C4SM1J. AWARE data are available via the ARM archive found at https://www.arm.gov/data/.
REFERENCES
Augustine, J. A., J. J. DeLuisi, and C. N. Long, 2000: SURFRAD—A national surface radiation budget network for atmospheric research. Bull. Amer. Meteor. Soc., 81, 2341–2358, https://doi.org/10.1175/1520-0477(2000)081<2341:SANSRB>2.3.CO;2.
Boeke, R. C., and P. C. Taylor, 2016: Evaluation of the Arctic surface radiation budget in CMIP5 models. J. Geophys. Res. Atmos., 121, 8525–8548, https://doi.org/10.1002/2016JD025099.
Collins, W. D., P. J. Rasch, B. E. Eaton, B. V. Khattatov, J.-F. Lamarque, and C. S. Zender, 2001: Simulating aerosols using a chemical transport model with assimilation of satellite aerosol retrievals: Methodology for INDOEX. J. Geophys. Res., 106, 7313–7336, https://doi.org/10.1029/2000JD900507.
Doelling, D. R., and Coauthors, 2013: Geostationary enhanced temporal interpolation for CERES flux products. J. Atmos. Oceanic Technol., 30, 1072–1090, https://doi.org/10.1175/JTECH-D-12-00136.1.
Doelling, D. R., M. Sun, L. T. Nguyen, M. L. Nordeen, C. O. Haney, D. F. Keyes, and P. E. Mlynczak, 2016: Advances in geostationary-derived longwave fluxes for the CERES synoptic (SYN1deg) product. J. Atmos. Oceanic Technol., 33, 503–521, https://doi.org/10.1175/JTECH-D-15-0147.1.
Dubovik, O., B. Holben, T. F. Eck, A. Smirnov, Y. J. Kaufman, M. D. King, D. Tanré, and I. Slutsker, 2002: Variability of absorption and optical properties of key aerosol types observed in worldwide locations. J. Atmos. Sci., 59, 590–608, https://doi.org/10.1175/1520-0469(2002)059<0590:VOAAOP>2.0.CO;2.
Fu, Q., and K. N. Liou, 1993: Parameterization of the radiative properties of cirrus clouds. J. Atmos. Sci., 50, 2008–2025, https://doi.org/10.1175/1520-0469(1993)050<2008:POTRPO>2.0.CO;2.
Gupta, S. K., 1989: A parameterization for longwave surface radiation from sun-synchronous satellite data. J. Climate, 2, 305–320, https://doi.org/10.1175/1520-0442(1989)002<0305:APFLSR>2.0.CO;2.
Gupta, S. K., W. L. Darnell, and A. C. Wilber, 1992: A parameterization for longwave surface radiation from satellite data: Recent improvements. J. Appl. Meteor. Climatol., 31, 1361–1367, https://doi.org/10.1175/1520-0450(1992)031<1361:APFLSR>2.0.CO;2.
Gupta, S. K., D. P. Kratz, P. W. Stackhouse Jr., and A. C. Wilber 2001: The Langley Parameterized Shortwave Algorithm (LPSA) for surface radiation budget studies. NASA Tech. Memo. NASA/TP-2001-211272, 31 pp.
Gupta, S. K., D. P. Kratz, P. W. Stackhouse Jr., A. C. Wilber, T. Zhang, and V. E. Sothcott, 2010: Improvement of surface longwave flux algorithms used in CERES processing. J. Appl. Meteor. Climatol., 49, 1579–1589, https://doi.org/10.1175/2010JAMC2463.1.
Ham, S.-H., S. Kato, and F. G. Rose, 2019: Impacts of partly cloudy pixels on shortwave broadband irradiance computations. J. Atmos. Oceanic Technol., 36, 369–386, https://doi.org/10.1175/JTECH-D-18-0153.1.
Ham, S.-H., S. Kato, F. G. Rose, S. Sun-Mack, Y. Chen, W. F. Miller, and R. C. Scott, 2022: Combining cloud properties from CALIPSO, CloudSat, and MODIS for top-of-atmosphere shortwave broadband irradiance computations: Impact of cloud vertical profiles. J. Appl. Meteor. Climatol., 61, 1449–1471, https://doi.org/10.1175/JAMC-D-21-0260.1.
Hess, M., P. Koepke, and I. Schult, 1998: Optical Properties of Aerosols and Clouds: The software package OPAC. Bull. Amer. Meteor. Soc., 79, 831–844, https://doi.org/10.1175/1520-0477(1998)079<0831:OPOAAC>2.0.CO;2.
Hu, Y., and K. Stamnes, 1993: An accurate parameterization of the radiative properties of water clouds suitable for use in climate models. J. Climate, 6, 728–742, https://doi.org/10.1175/1520-0442(1993)006<0728:AAPOTR>2.0.CO;2.
Huang, G., and Coauthors, 2020: The accuracy improvement of clear-sky surface shortwave radiation derived from CERES SSF dataset with a simulation analysis. Sci. Total Environ., 749, 141671, https://doi.org/10.1016/j.scitotenv.2020.141671.
Ignatov, A., and Coauthors, 2005: Two MODIS aerosol products over ocean on the Terra and Aqua CERES SSF datasets. J. Atmos. Sci., 62, 1008–1031, https://doi.org/10.1175/JAS3383.1.
Inamdar, A., and V. Ramanathan, 1997: On monitoring the atmospheric greenhouse effect from space. Tellus, 49B, 216–230, https://doi.org/10.3402/tellusb.v49i2.15963.
Jin, Z., T. P. Charlock, W. L. Smith Jr., and K. Rutledge, 2004: A parameterization of ocean surface albedo. Geophys. Res. Lett., 31, L22301, https://doi.org/10.1029/2004GL021180.
Kato, S., T. P. Ackerman, J. H. Mather, and E. E. Clothiaux, 1999: The k-distribution method and correlated-k approximation for a shortwave radiative transfer model. J. Quant. Spectrosc. Radiat. Transfer, 62, 109–121, https://doi.org/10.1016/S0022-4073(98)00075-2.
Kato, S., F. G. Rose, and T. P. Charlock, 2005: Computation of domain-averaged irradiance using satellite-derived cloud properties. J. Atmos. Oceanic Technol., 22, 146–164, https://doi.org/10.1175/JTECH-1694.1.
Kato, S., and Coauthors, 2018: Surface irradiances of Edition 4.0 Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) data product. J. Climate, 31, 4501–4527, https://doi.org/10.1175/JCLI-D-17-0523.1.
Kratz, D. P., and F. G. Rose, 1999: Accounting for molecular absorption within the spectral range of the CERES window channel. J. Quant. Spectrosc. Radiat. Transfer, 61, 83–95, https://doi.org/10.1016/S0022-4073(97)00203-3.
Kratz, D. P., S. K. Gupta, A. C. Wilber, and V. E. Sothcott, 2010: Validation of the CERES Edition 2B surface-only flux algorithms. J. Appl. Meteor. Climatol., 49, 164–180, https://doi.org/10.1175/2009JAMC2246.1.
Kratz, D. P., P. W. Stackhouse Jr., S. K. Gupta, A. C. Wilber, P. Sawaengphokhai, and G. R. McGarragh, 2014: The Fast Longwave and Shortwave Flux (FLASHFlux) data product: Single-scanner footprint fluxes. J. Appl. Meteor. Climatol., 53, 1059–1079, https://doi.org/10.1175/JAMC-D-13-061.1.
Kratz, D. P., S. K. Gupta, A. C. Wilber, and V. E. Sothcott, 2020: Validation of the CERES Edition-4A surface-only flux algorithms. J. Appl. Meteor. Climatol., 59, 281–295, https://doi.org/10.1175/JAMC-D-19-0068.1.
Levy, R. C., S. Mattoo, L. A. Munchak, L. A. Remer, A. M. Sayer, F. Patadia, and N. C. Hsu, 2013: The Collection 6 MODIS aerosol products over land and ocean. Atmos. Meas. Tech., 6, 2989–3034, https://doi.org/10.5194/amt-6-2989-2013.
Li, Z., and L. Garand, 1994: Estimation of surface albedo from space: A parameterization for global application. J. Geophys. Res., 99, 8335–8350, https://doi.org/10.1029/94JD00225.
Loeb, N. G., K. J. Priestley, D. P. Kratz, E. B. Geier, R. N. Green, B. A. Wielicki, P. O. Hinton, and S. K. Nolan, 2001: Determination of unfiltered radiances from the Clouds and the Earth’s Radiant Energy System instrument. J. Appl. Meteor. Climatol., 40, 822–835, https://doi.org/10.1175/1520-0450(2001)040<0822:DOURFT>2.0.CO;2.
Loeb, N. G., N. Manalo-Smith, W. Su, M. Shankar, and S. Thomas, 2016: CERES top-of-atmosphere Earth radiation budget climate data record: Accounting for in-orbit changes in instrument calibration. Remote Sens., 8, 182, https://doi.org/10.3390/rs8030182.
Loeb, N. G., and Coauthors, 2018a: Clouds and the Earth’s Radiant Energy System (CERES) Energy Balanced and Filled (EBAF) top-of-atmosphere (TOA) Edition-4.0 data product. J. Climate, 31, 895–918, https://doi.org/10.1175/JCLI-D-17-0208.1.
Loeb, N. G., and Coauthors, 2018b: Impact of ice cloud microphysics on satellite cloud retrievals and broadband flux radiative transfer model calculations. J. Climate, 31, 1851–1864, https://doi.org/10.1175/JCLI-D-17-0426.1.
Lubin, D., and Coauthors, 2020: AWARE: The Atmospheric Radiation Measurement (ARM) West Antarctic Radiation Experiment. Bull. Amer. Meteor. Soc., 101, E1069–E1091, https://doi.org/10.1175/BAMS-D-18-0278.1.
Lucchesi, R., 2013: File specification for GEOS-5 FP-IT (Forward Processing for Instrument Teams). GMAO Office Tech. Rep. 2, 63 pp.
Minnis, P., and Coauthors, 2020: CERES MODIS cloud product retrievals for Edition 4—Part I: Algorithm changes. IEEE Trans. Geosci. Remote Sens., 59, 2744–2780, https://doi.org/10.1109/TGRS.2020.3008866.
Nicolas, J. P., and Coauthors, 2017: January 2016 extensive summer melt in West Antarctica favoured by strong El Niño. Nat. Commun., 8, 15799, https://doi.org/10.1038/ncomms15799.
Ohmura, A., and Coauthors, 1998: Baseline Surface Radiation Network (BSRN/WCRP): New precision radiometry for climate research. Bull. Amer. Meteor. Soc., 79, 2115–2136, https://doi.org/10.1175/1520-0477(1998)079<2115:BSRNBW>2.0.CO;2.
Oreopoulos, L., and Coauthors, 2012: The Continual Intercomparison of Radiation Codes: Results from phase I. J. Geophys. Res., 117, D06118, https://doi.org/10.1029/2011JD016821.
Remer, L. A., and Coauthors, 2005: The MODIS aerosol algorithm, products, and validation. J. Atmos. Sci., 62, 947–973, https://doi.org/10.1175/JAS3385.1.
Rienecker, M. M., and Coauthors, 2008: The GEOS-5 data assimilation system: Documentation of versions 5.0.1, 5.1.0, and 5.2.0. NASA Tech. Memo. NASA/TM-2008-104606, Vol. 27, 118 pp.
Rose, F. G., D. A. Rutan, T. Charlock, G. L. Smith, and S. Kato, 2013: An algorithm for the constraining of radiative transfer calculations to CERES-observed broadband top-of-atmosphere irradiance. J. Atmos. Oceanic Technol., 30, 1091–1106, https://doi.org/10.1175/JTECH-D-12-00058.1.
Rutan, D. A., T. P. Charlock, F. Rose, S. Kato, S. M. Zentz, and L. H. Coleman, 2006: Global surface albedo from CERES/Terra Surface and Atmospheric Radiation Budget (SARB) data product. 12th Conf. on Atmospheric Radiation, Madison, WI, Amer. Meteor. Soc., P1.1, https://ams.confex.com/ams/Madison2006/techprogram/paper_112637.htm.
Rutan, D. A., F. Rose, M. Roman, N. Manalo-Smith, C. Schaaf, and T. Charlock, 2009: Development and assessment of broadband surface albedo from Clouds and the Earth’s Radiant Energy System Clouds and Radiation Swath data product. J. Geophys. Res., 114, D08125, https://doi.org/10.1029/2008JD010669.
Rutan, D. A., S. Kato, D. R. Doelling, F. G. Rose, L. T. Nguyen, T. E. Caldwell, and N. G. Loeb, 2015: CERES synoptic product: Methodology and validation of surface radiant flux. J. Atmos. Oceanic Technol., 32, 1121–1143, https://doi.org/10.1175/JTECH-D-14-00165.1.
Schaaf, C. B., and Coauthors, 2002: First operational BRDF, albedo nadir reflectance products from MODIS. Remote Sens. Environ., 83, 135–148, https://doi.org/10.1016/S0034-4257(02)00091-3.
Scott, R. C., J. P. Nicolas, D. H. Bromwich, J. R. Norris, and D. Lubin, 2019: Meteorological drivers and large-scale climate forcing of West Antarctic surface melt. J. Climate, 32, 665–684, https://doi.org/10.1175/JCLI-D-18-0233.1.
Shupe, M. D., and Coauthors, 2022: Overview of the MOSAiC expedition—Atmosphere. Elementa, 10, 00060, https://doi.org/10.1525/elementa.2021.00060.
Stokes, G. M., and S. E. Schwartz, 1994: The Atmospheric Radiation Measurement (ARM) program: Programmatic background and design of the Cloud and Radiation Test Bed. Bull. Amer. Meteor. Soc., 75, 1201–1222, https://doi.org/10.1175/1520-0477(1994)075<1201:TARMPP>2.0.CO;2.
Su, W., T. P. Charlock, and F. G. Rose, 2005: Deriving surface ultraviolet radiation from CERES surface and atmospheric radiation budget: Methodology. J. Geophys. Res., 110, D14209, https://doi.org/10.1029/2005JD005794.
Su, W., T. P. Charlock, F. G. Rose, and D. Rutan, 2007: Photosynthetically active radiation from Clouds and the Earth’s Radiant Energy System (CERES) products. J. Geophys. Res., 112, G02022, https://doi.org/10.1029/2006JG000290.
Su, W., J. Corbett, Z. Eitzen, and L. Liang, 2015a: Next-generation angular distribution models for top-of-atmosphere radiative flux calculation from CERES instruments: Methodology. Atmos. Meas. Tech., 8, 611–632, https://doi.org/10.5194/amt-8-611-2015.
Su, W., J. Corbett, Z. Eitzen, and L. Liang, 2015b: Next-generation angular distribution models for top-of-atmosphere radiative flux calculation from CERES instruments: Validation. Atmos. Meas. Tech., 8, 3297–3313, https://doi.org/10.5194/amt-8-3297-2015.
Su, W., L. Liang, H. Wang, and Z. A. Eitzen, 2020: Uncertainties in CERES top-of-atmosphere fluxes caused by changes in accompanying imager. Remote Sens., 12, 2040, https://doi.org/10.3390/rs12122040.
Tegen, I., and A. A. Lacis, 1996: Modeling of particle size distribution and its influence on the radiative properties of mineral dust aerosol. J. Geophys. Res., 101, 19 237–19 244, https://doi.org/10.1029/95JD03610.
Wilson, A., R. C. Scott, M. P. Cadeddu, V. Ghate, and D. Lubin, 2018: Cloud optical properties over West Antarctica from shortwave spectroradiometer measurements during AWARE. J. Geophys. Res. Atmos., 123, 9559–9570, https://doi.org/10.1029/2018JD028347.
Yost, C. R., P. Minnis, S. Sun-Mack, Y. Chen, and W. L. Smith, 2020: CERES MODIS cloud product retrievals for Edition 4—Part II: Comparisons to CloudSat and CALIPSO. IEEE Trans. Geosci. Remote Sens., 59, 3695–3724, https://doi.org/10.1109/TGRS.2020.3015155.
Zhou, Y., and R. D. Cess, 2001: Algorithm development strategies for retrieving the downwelling longwave flux at the Earth’s surface. J. Geophys. Res., 106, 12 477–12 488, https://doi.org/10.1029/2001JD900144.
Zhou, Y., D. P. Kratz, A. C. Wilber, S. K. Gupta, and R. D. Cess, 2007: An improved algorithm for retrieving surface downwelling longwave radiation from satellite measurements. J. Geophys. Res., 112, D15102, https://doi.org/10.1029/2006JD008159.