The CERES Cloud Radiative Swath: Enhanced Level 2 Surface Fluxes with the Fu-Liou RTM & Machine Learning
Explore the source record for details and available documents.
Ryan C Scott↗
Engineering topics
Patrick C Taylor
Publications and source records attributed to Patrick C Taylor.
Isolating the Surface Type Influence on Arctic Cloud Properties
In response to anthropogenic climate change, substantial declines in sea ice have been observed. An urgent question is whether and how clouds respond to the changing Arctic surface. Due to the important radiative influence of clouds, the response of clouds to sea ice cover change constitutes a potentially important climate feedback within the Arctic. The sign of this feedback is strongly influenced by the seasonality of the cloud response to sea ice loss. Given the importance of this phenomenon, previous research focused on using observations of inter-annual variability to quantify the cloud-sea ice relationship. However, the reliance on inter-annual variability to assess the covariance between clouds and sea ice make it challenging to control for the influence of large-scale meteorology on the clouds. We have devised an approach that applies an in situ, airborne observational strategy to quantify the influence of surface type on Arctic clouds to satellite data. Our event-based method composites cloud property differences for ocean, marginal ice zone, and sea ice surface types for >5000 MIZ crossing events. We find that cloud properties in non-summer months below ~1.5 km are influenced by surface type such that greater cloud fraction and water content are found over ocean relative to sea ice regions. Results are statistically significantly at the 95% confidence level. During summer, the surface type does not influence the cloud properties. The primary cause of the surface type dependent cloud property differences is the thermodynamic profiles differences that occur between the two surface types: namely, ocean footprints are warmer, moister, have more positive surface turbulent fluxes and are less stable than their sea ice counterparts. We find that the differences in thermodynamic profiles by surface type that correspond to the cloud property differences also explains a significant portion of the variability across events. Our analysis provides evidence that the influence of surface type on cloud properties fundamentally occurs due to the inherent surface temperature differences between the surface types. Significant surface type influences on cloud properties are only found in the presence of surface temperature differences between the surface types and only non-summer months meet this criterion. A conceptual model of this process is as follows, fundamental properties of the ocean and sea ice surface types (e.g., surface albedo, heat capacity/thermal inertia, surface roughness, and thermal conductance) lead to systematic differences in the surface temperature by surface type. In response to this surface temperature difference, atmosphere boundary layer processes (including turbulent mixing and radiation) communicate the surface temperature differences to the lower atmosphere driving the greater stability over sea ice than ocean. These differences in the background thermodynamic state profiles between the surface types yield atmospheric conditions that are more conducive to cloud development over ocean than sea ice. The details of these boundary layer processes are sensitive to the surface type but may not be fundamental to the resulting thermodynamic structure and cloud differences. Thus, our results indicate that to accurately model that cloud response to sea ice loss in climate models the surface type properties that yield the systematic differences in ocean and sea ice temperature (e.g., surface albedo, surface turbulent fluxes parameters, and thermal conductance) must be accurately represented.
Arctic radiation budget↗
Radiative Flux Measurements from ARISE: A Comparison with CERES Top-of-Atmosphere Radiative Fluxes
Uncertainty in top-of-atmosphere (TOA) radiation fluxes observations are larger in the Arctic than in other regions. These uncertainties are due to the low sun angles and the highly reflective, anisotropic, and heterogeneous surface conditions. Quantifying, attributing, and reducing Arctic TOA radiative flux uncertainty enables a better understanding of the rapidly changing Arctic. To advance this goal, we compare the Cloud and Earth’s Radiant Energy System (CERES) TOA radiative fluxes with Arctic Radiation-IceBridge Sea and Ice Experiment (ARISE) campaign measurements collected in September 2014. We compare CERES TOA and aircraft radiative flux measurements using two complementary approaches: grid box average fluxes and instantaneously matched footprints. The grid box mean flux comparison indicates an agreement between CERES and aircraft measurements within 2 uncertainty (calibration and inversion) in the longwave for all five grid boxes and for four-of-five grid boxes in the shortwave; shortwave and longwave mean differences are -7.9 and +2.3 Wm‑2, respectively. The comparison of 36 instantaneously matched footprints with aircraft measurements reveals mean differences of -10.5 and 0.4 Wm‑2 in the shortwave and longwave, respectively. To further explore the persistent negative difference in the shortwave, we further quantify the effects of temporal and spatial sampling differences, angular distribution models, and scene identification to CERES-aircraft differences. Our analysis indicates that sampling differences (including scene evolution) account for an additional 1.8 and 1.7% uncertainty in the shortwave and longwave, respectively and indicates no bias. After accounting for this sampling uncertainty, all CERES-aircraft grid box mean fluxes agree within 2 uncertainty. Scene identification errors due to sea ice concentration data set differences exhibit no bias in the shortwave flux difference and indicate the possibility of substantial differences in the CERES fluxes in specific cases with large spatial heterogeneity. Considering the instantaneously matched footprints, we find that the angular distribution models account may account for up to ‑7.3 Wm-2 of the persistent CERES-aircraft shortwave flux difference due to systematic differences in the anisotropy for sea ice partly cloudy scenes. Additional analysis using a special scan model with the CERES FM2 instrument suggests a significant view zenith angle dependence of the CERES fluxes for sea ice partly cloudy scenes where shortwave fluxes systematically decrease with increasing view zenith angle; no dependence is found for other scene types. We conclude that (1) spatial heterogeneity and scene temporal evolution substantially limit our ability to use aircraft measurements to place strong constraints on CERES TOA fluxes and (2) that the representation of anisotropy in sea ice partly cloudy scenes is a significant factor contributing to the persistent negative CERES-aircraft shortwave flux difference in this comparison and require additional data to analysis fully quantify the potential bias.
Patrick C Taylor↗
Arctic Amplification: Process Drivers and Sources of Uncertainty
Arctic amplification (AA) is a coupled atmosphere-sea ice-ocean process. This understanding has evolved from the early concept of AA, as a consequence of snow ice line progressions, through more than a century of research that has clarified the relevant processes and driving mechanisms of AA. The predictions made by early modeling studies, namely the fall/winter maximum, bottom-heavy structure, the prominence of surface albedo feedback, and the importance of stable stratification have withstood the scrutiny of multi-decadal observations and more complex models. Yet, the uncertainty in Arctic climate projections is larger than in any other region of the planet, making the assessment of high-impact, near-term regional changes difficult or impossible. Reducing this large spread in Arctic climate projections requires a quantitative process understanding. This presentation synthesizes current knowledge of AA and describes a set of recommendations to guide future research. It briefly reviews the history of AA science, summarizes observed Arctic changes, discusses modeling approaches and feedback diagnostics, and assesses the current understanding of the most relevant feedbacks to AA. These sections culminate in a conceptual model of the fundamental physical mechanisms causing AA and a collection of recommendations to accelerate progress towards reduced uncertainty in Arctic climate projections. Our conceptual model highlights the need to account for local feedback and remote process interactions within the context of the annual cycle to constrain projected AA.
Patrick C Taylor↗
CERES Top-of-atmosphere and Surface fluxes in the Arctic: A comparison with ARISE and MOSAiC measurements
Uncertainty in the NASA Clouds and the Earth’s Radiant Energy System (CERES)-derived irradiances is larger over sea ice than any other surface type and comes from several sources. This presentation summarizes results from comparisons of CERES data with the year-long Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition in the central Arctic and the Arctic Radiation and Icebridge Sea ice Experiment (ARISE). The CERES Synoptic 1degree (SYN1deg) product overestimates the downwelling shortwave flux by 11.40 Wm–2 and underestimates the upwelling shortwave flux by –15.70 Wm–2 and downwelling longwave fluxes by –12.58 Wm–2 at the surface during summer. In addition, large differences are found in the upwelling longwave flux when the surface approaches the melting point (approximately 0C). The large negative bias in upwelling shortwave flux can be attributed in large part to lower surface albedo (–0.15) in satellite footprint relative to surface sensors. The CERES-MOSAiC broadband albedo differences (approximately 20 Wm–2) explain a larger portion of the upwelling shortwave flux difference than the spectral albedo shape differences (approximately 3 Wm–2). The ARISE results indicate the some of these differences in the SW are due to the scene identification and representation of anisotropy in partly cloud scenes. Switching from imager-based to passive microwave-based sea ice data in the CERES inversion process reduces the differences in the grid box average fluxes and in the sea ice partly cloudy scene anisotropy in the instantaneously-matched footprints. Our analysis indicates that calibration and sampling uncertainty limit the ability to place strong constraints (<7%) on CERES TOA fluxes with aircraft measurements. The results indicate that improvements in the surface albedo and cloud data would substantially reduce the uncertainty in the Arctic surface radiation budget derived from CERES data products.
Patrick C Taylor↗
An Observationally-Based Determination of the Arctic Sea Ice-Cloud Feedback Since 2000: Isolating the Arctic Cloud Response to Sea Ice Loss
Arctic sea ice responds to and drives Arctic climate change. The interactions between Arctic sea ice and clouds represent a mechanism through which sea ice can drive climate change. We composite active remote sensing satellite cloud properties for ice-free, marginal ice zone (MIZ), and ice-covered surfaces during MIZ crossing events to investigate the influence of the transition from an ice-covered to an ice-free surface on low-level clouds. We demonstrate that the event-based methodology controls for large-scale meteorological factors and isolates the sea ice effect on clouds. We find larger cloud fraction and total water content below ~1.5 km over ice-free relative to ice-covered surfaces during non-summer months, indicating a low-level cloud sensitivity to Arctic sea ice decline. During summer, results show larger cloud fraction and water content over ice-free surfaces, however the differences are statistically indistinguishable. Evidence is provided that atmospheric thermodynamic profile differences cause the cloud property differences, namely that ice-free footprints are warmer, moister, have more positive surface turbulent fluxes and are less stable than their ice-covered counterparts. Ice-free and ice-covered surface cloud property differences scale with surface temperature differences such that cloud property differences are only found in the presence of a surface temperature difference. We conclude that surface temperature differences modulate the cloud response to sea ice loss through influences on surface turbulent fluxes and lower tropospheric stability. The results imply a positive non-summer sea ice-cloud feedback and that up to a 0.02 cloud fraction and 0.05 g m 3 total water content increase in fall are due to the observed sea ice decline.
Patrick C Taylor↗
The Importance of Cross-Scale and Cross-Interface Processes on Arctic Amplification
The Arctic is a dynamic region, demonstrated by its remarkable internal variability and rapid response to anthropogenic climate change over the last >40 years. Actionable predictions and projections hold significant value for managing both natural and human systems. The value of these model outputs only grows in a warmer, less icy Arctic. However, substantial gaps exist in our understanding of cross-scale and cross-interface (air-sea ice-ocean) interactions that limit our predictive capabilities. Studies dating back at least 40 years provide evidence that the evolution of the Arctic climate system is sensitive to these cross-scale and cross-interface interactions. This presentation summarizes our understanding of cross-scale and cross-interface interactions relevant to the Arctic’s response to anthropogenic climate change—Arctic Amplification. The presentation emphasizes the influence of surface-type dependent turbulent flux exchanges of heat and moisture, the rectification of episodic atmospheric heat transport events on time-averaged changes, local and remote feedback interactions, and cross-seasonal energy transfers. The presentation concludes with a summary of knowledge gaps and discusses potential pathways for accelerating our understanding of the Arctic climate system.
Patrick C Taylor↗
Towards a More Realistic Representation of NASA CERES-derived Surface Radiative Fluxes during Polar Night: A Comparison with the MOSAiC Field Campaign
The Arctic is one of the most sensitive regions of Earth to climate change, and yet remains one of the more difficult regions to both observe and simulate. The remoteness of the Arctic from most of civilization makes large volumes of in situ observations difficult to collect, and so satellite observations are a critical tool for observing the region, such as those from Clouds and the Earth’s Radiant Energy System (CERES). But validating the satellite radiative flux estimates is difficult because of the lack of in situ measurements. A useful remedy is utilizing measurements from field campaigns, such as the year-long Multidisciplinary drifting Observatory for the Study of Arctic Climate (MOSAiC) expedition during 2019-2020. The extensive high-quality surface radiative flux and meteorological measurements collected from MOSAiC provide a useful check on flux retrievals from CERES instruments. We compare MOSAiC and CERES surface radiative fluxes during October through March of 2019-2020. As this time period is mostly during polar night and twilight, the focus of the study is on longwave surface fluxes. To better understand the reasons for errors in CERES fluxes, we use the large set of meteorological measurements also collected by MOSAiC, including surface temperature, thermodynamic vertical profiles, surface turbulent fluxes, and cloud properties. Previous work identified a significant source of error in the CERES estimate of surface downwelling longwave flux as the estimate of low level cloud amount when compared with MOSAiC W-band radar measurements. An overestimate of low cloud amount yielded an underestimate of the downwelling flux on the order of tens of W m2. Continuing this work, we test other meteorological properties such as the vertical thermodynamic profile and surface turbulent fluxes. In addition, we examine different estimates of surface fluxes derived from CERES measurements. Previous work used the SYN1deg product, which had the limitation of having a 1º x 1 º horizontal resolution. A newer product estimates the surface fluxes at the CERES footprint resolution, which allows a more precise colocation, and thus a representative flux estimate, with the MOSAiC site.
J Brant Dodson↗
Quantifying Changes in the Arctic Shortwave Cloud Radiative Effects
The shortwave cloud radiative effect (SWCRE) is important on the Arctic surface radiation budget and the major source of inter-model spread in predictions of Arctic climate. To better understand the individual contributions of various radiative processes to changes in SWCRE, the paper presents the use of the extended APRP (Atmospheric Radiative Perturbation Potential) method. This involves adding the absorptivity for the upward beam and considering differences in reflectivity between upward and downward beams, as well as analyzing the cloud masking effect resulting from changes in surface albedo in more detail. Using data from the CMIP5 and CMIP6 climate models, the study decomposes the SWCRE over the Arctic surface and analyzes inter-model differences in quadrupled CO2 simulations. The study takes into account the fact that the response of SWCRE to Arctic warming is influenced by changes in surface albedo, cloud amount, and cloud microphysics. Results show that in the sunlight season, the reduction in surface albedo associated with sea ice loss is directly linked to strong negative SWCRE, which explains the considerable model discrepancy. Arctic clouds can hinder the positive surface albedo feedback by changing the albedo in two ways: (1) decreasing incoming shortwave radiation due to cloud reflection and (2) by decreasing the shortwave reaching the surface after being reflected by clouds. In addition, increased (decreased) cloud amount and cloud liquid water are shown to be less (more) incoming shortwave fluxes at the surface, but not dominating factors to the Arctic surface radiation budget and its inter-model variation. Overall, the extended APRP method offers a useful tool for analyzing the complex interactions between clouds and radiative process, reasonably decomposes the individual SWCRE responses at the Arctic surface, and emphasizes that considering not only the cloud amount or its properties, but also surface albedo change is critical for the prediction of SWCRE on the Arctic surface.
Arctic cloud↗
Towards a More Realistic Representation of NASA CERES-derived Surface Radiative Fluxes during Polar Night: A Comparison with the MOSAiC Field Campaign
The Arctic remains one of the more difficult regions observe, and so satellite observations are a critical tool for observing the region, such as those from Clouds and the Earth’s Radiant Energy System (CERES). But validating the satellite surface radiative flux estimates is difficult because of the lack of in situ measurements. The extensive high-quality surface radiative flux and meteorological measurements collected from MOSAiC provide a useful check on flux retrievals from CERES instruments. We compare MOSAiC and CERES surface radiative fluxes during October through March of 2019-2020 using the large set of meteorological measurements also collected by MOSAiC, specifically cloud properties. Previous work identified a significant source of error in the CERES estimate of surface downwelling longwave flux as the estimate of low level cloud amount when compared with MOSAiC W-band radar measurements. Continuing this work, we examine the effects of errors in cloud water path on surface radiative fluxes. When using all cloud conditions, errors in cloud water are also significantly correlated with surface radiative flux errors, though the size off the effect is only about half that of cloud amount. But when considering low cloud conditions only, the effects of cloud water and cloud amount are comparable. We compare MOSAiC and CERES surface radiative fluxes during October through March of 2019-2020 using the large set of meteorological measurements also collected by MOSAiC, specifically cloud properties. Previous work identified a significant source of error in the CERES estimate of surface downwelling longwave flux as the estimate of low level cloud amount when compared with MOSAiC W-band radar measurements. We further examine this source of error by examining selected case studies in which the disagreements in radiative fluxes and clouds are large.
J Brant Dodson↗
The Influence of Parameterization Schemes on Arctic Low Cloud Properties and their Variability in HadGEM3
Explore the source record for details and available documents.
Robyn C Boeke↗
Clouds in the Arctic Climate System: A Force for Change?
Clouds play a central role in the global climate system through the modulation of Earth’s energy flows and as a mediator of precipitation. In the Arctic, clouds are also a major player; however, the processes that govern cloud evolution in the Arctic differ from most of the other regions of the globe. Thus, clouds are a key “wildcard” within the Arctic climate system that could have a substantial influence on the Arctic climate system response to anthropogenic forcing. What is the role of clouds within the phenomenon known as Arctic Amplification? This is question does not have a clear answer. Clouds seem to be a center to the important processes driving Arctic Amplification (e.g., the atmospheric response to sea ice loss and airmass transformation), however feedback analysis studies indicate that net cloud feedback in the Arctic is small. This seminar discusses the role of clouds within the Arctic Amplification processes reviewing aspects of what we know about Arctic clouds and the uncertainties that limit our ability to model them. Results from recently published and ongoing work are presented that provide an observationally-based estimate of the cloud-sea ice feedback and evaluate cloud properties within models. The goal of this presentation is to ignite a discussion and new collaborations around the best approaches to resolving uncertainties related to the role of clouds with the Arctic system and how to better represent Arctic clouds in models.
Patrick C Taylor↗
The Surface Albedo of Sea Ice in CMIP6 and the Implications for the Surface Albedo Feedback
The Arctic has experienced rapid sea ice loss and a substantial decline in surface albedo, significantly impacting its radiation budget. Climate models from the Coupled Model Intercomparison Project (CMIP6) reproduce these changes. However, inconsistencies remain among models regarding the magnitude, spatial distribution, and seasonal patterns of Arctic surface albedo evolution. This study investigates these discrepancies by comparing model outputs with observation from the Clouds and the Earth's Radiant Energy System (CERES). We develop a decomposition method to assess the contributions of sea ice albedo, sea ice concentration, and sea ice extent to Arctic surface albedo. Over land, differences in snow cover account for the substantial inter-model spread in surface albedo, while over the ocean, sea ice albedo, concentration, and extent all contribute. Comparisons between CMIP6 and the Atmospheric Model Intercomparison Project (AMIP) simulations, which use identically prescribed sea ice concentrations, reveals considerable inter-model spread in Arctic Ocean surface albedo due to differences in sea ice albedo. Applying the decomposition method to projections shows that models predicting larger decreases in sea ice concentration and extent, especially in the Central Arctic, exhibit lower surface albedo and stronger sea ice albedo feedback. Beyond 2050, Arctic Ocean surface albedo decline is mainly influenced by sea ice extent indicating that the retreat of the ice edge is the most important process to constrain the surface albedo feedback. This study provides insights into factors contributing to the spread and changes in Arctic surface albedo and the associated sea ice albedo feedback.
Patrick C Taylor↗
A Comparative Analysis of the Biases and Trends in NEX GDDP Downscaled Model Data at Different NASA Centers and Facilities GC13F-0245
Explore the source record for details and available documents.
Bradley M Hegyi↗
Are You In or Out: the Influence of Sea Ice Drift on Sea Ice Survivability
The state of Arctic sea ice influences aspects of many global systems including ecosystems, economies, geopolitics, and climate. And, it is in severe decline. Numerous studies have considered the factors that influence Arctic sea ice decline using monthly mean or gridded data. However, these methods struggle to cleanly separate the influences on sea ice melt because the sea ice cover is always moving; thus, different sea ice floes and sea ice regimes constitute the monthly average sea ice state within the Eulerian frame. This study uses a LaGrangian sea ice parcel tracking satellite database to investigate the factors that influence sea ice parcel survivability. The concept of sea ice survivability is defined as the likelihood that a sea ice parcel will last through the summer melt season. Survivability can be stratified in many ways to analyze how it changes as a function of sea ice regime, region, and sea ice characteristics. How is the survivability of first year (FY) and multi-year (MY) ice classes influenced by sea ice motion? This is the central question addressed in this presentation. Specifically, this analysis determines sea ice survivability as a function of distance traveled by sea ice parcels and compares the survivability statistics for sea ice parcels that stay in their starting region versus those that leave their starting region. The preliminary results indicate that the influence of sea ice movement on survivability depends strongly on which region the sea ice resides in at the beginning of the growth season. Lastly, we find substantial inter-annual variability in the regional distribution of sea ice survivability that seems to be strongly linked to sea ice parcel drift between regions of the Arctic Ocean. Moreover, we use these results to discuss the potential contribution of changes in sea ice motion on the observed slowdown in Arctic sea ice extent loss.
Patrick C Taylor↗
A Path to Improving Simulated Properties of Low Clouds over the Beaufort Sea using Airborne In Situ Observations of Subgrid-Scale Variability
Arctic low clouds influence the evolution of the Arctic system through their effects on radiative fluxes, boundary layer mixing, stability, turbulence, humidity, and precipitation. Unfortunately, atmospheric models and retrospective analysis (reanalysis) products struggle to accurately simulate the occurrence and properties of low clouds in the Arctic. One of the main reasons for this problem are the possible unrealistic assumptions that models/reanalyses make about the subgrid-scale (SGS) variability of meteorological properties, as well as the relationship between SGS variability and grid-scale (GS) cloud properties. We utilize cloud and thermodynamic data of low level (primarily) liquid clouds collected from two aircraft campaigns conducted over the Beaufort Sea to better understand and characterize this problem. Examining data from the September 2014 Arctic Radiation-IceBridge Sea and Ice Experiment (ARISE) airborne campaign and the 1998 First International Satellite Cloud Climatology Project (ISCCP) Regional Experiment (FIRE)–Arctic Cloud Experiment (ACE) reveals that GS cloud water variability is closely related with SGS distribution of total water (i.e. water vapor + cloud water). We examine the influence of three assumed SGS parameters on the estimation of GS cloud water: the width of the SGS PDF of total water (ΔQT), the shape of the SGS PDF, and the critical saturation ratio (Scrit), which partitions the SGS PDF into water that is available or not for conversion to cloud water. Both Scrit and ΔQT influence predicted GS cloud water strongly. ARISE and FIRE-ACE disagree somewhat in the details, but both campaigns support the possibility that more realistic representations of these two parameters may lead to more realistic GS cloud water. In particular, both parameters are sensitive to GS relative humidity, and so we examine the effects of allowing these quantities to vary as an empirically-derived linear function of GS relative humidity. In contrast, the shape of the PDF has little effect on the predicted cloud water – a surprising result that merits additional investigation.
J Brant Dodson↗