BIG PICTURE CONTEXT: THE ROLE OF ARCTIC SEA ICE LOSS VS. OTHER FORCING
Introduction to Discussion of Big Picture Context
This talk raised some general points about the issues of distinguishing Arctic sea ice loss from other forcings in considering mid-latitude weather and extremes in preparation for a broad brainstorming session in the first morning of the workshop. We started with the idea that Arctic sea ice loss is in a sense the middle of a story that really begins with amplified Arctic change forced by greenhouse warming. In this bigger picture, Arctic sea ice, atmosphere, and ocean processes are strongly coupled and are all relevant controls on mid-latitude weather. We encouraged participants to think systematically about how Arctic changes might directly influence the mid-latitudes, and about how changes in mid-latitudes might be affected by the Arctic through feedbacks. We also pointed out that any change that projects onto modes of variability such as the NAM/AO may simultaneously affect the Arctic and mid-latitudes, making physical attribution more difficult. Finally, we proposed that efforts should be made to be precise about the timescale and seasonality of any proposed linkages, to think about the quantitative evidence for our current understanding (such as the ACID test of Screen), and to consider critical research gaps (observational/modelling/theoretical/impacts-related) in addressing the important question of Arctic influence on mid-latitude weather and extremes.
The Role of Arctic Sea Ice Loss James Screen, University of Exeter
Many linkages have been proposed between Arctic climate change and mid-latitude weather. These have received substantial media coverage. In a recent New Hampshire poll, 60 percent of respondents answered that they thought that future rapid Arctic warming would have major effects on the weather where they live. This talk reviewed published scientific evidence for proposed Arctic-mid-latitude linkages, including between Arctic sea ice loss and cold mid-latitude winters, and the evidence for, and against, changing planetary wave amplitudes in response to greater warming in the Arctic than at lower latitudes. A framework for assessing confidence in proposed linkages was introduced, the “ACID” test. It was argued that to have high confidence in a linkage it should be Attributable to Arctic forcing, Corroborated by multiple lines of evidence, Informed by mechanistic understanding and Detectable in the real world. It was argued that few, if any, of the proposed linkages currently pass the ACID test. A number of key uncertainties were highlighted, including disagreement between the atmospheric responses to sea ice loss in
different modeling studies. Also, it discussed how responses in some variables and locations are easier to detect (i.e. are stronger) than others. Pertinently, mid-latitude atmospheric responses are hard to detect due to remoteness from the sea-ice forcing (changes in surface fluxes) and large atmospheric internal variability. Thus, a degree of caution is required when linking sea ice loss to recent extreme weather events and possible trends therein. In short, the science behind Arctic-mid-latitude linkages is still in its infancy and more research is required to make more confident statements.
Present-Day and Future Projections of Mid-latitude Variability Elizabeth Barnes, Colorado State University
Mid-latitude atmospheric variability is composed of complex dynamical interactions between circulations of a range of spatial and temporal scales. For example, the low-frequency variability of the Northern Hemisphere jet-stream is tightly coupled to the presence and absence of high-frequency patterns such as blocking anticyclones and Rossby wave-breaking.
It has been hypothesized that Arctic amplification has increased the variability of the jet-stream and the frequency of blocking over the North Atlantic/North America region, however, no statistically significant increases in blocking frequency have been observed. It is suggested that given the large internal variability of the jet-stream and blocking, any potential effect of Arctic amplification on mid-latitude weather is unlikely to be detectable with current observations.
Looking to the future, the model simulations performed for CMIP5 (the Coupled Model Intercomparison Project, phase 5) offer mixed evidence of a link between Arctic amplification and mid-latitude circulation patterns. Blocking frequencies are projected to decrease over the oceans with climate change, opposite to the increase hypothesized to go with Arctic amplification. However, significant correlations between atmospheric wave properties and Arctic temperatures are shown. Idealized model experiments are highlighted as a useful way forward to test proposed dynamical mechanisms linking Artic amplification and mid-latitude variability.
Increasing greenhouse gas concentrations are projected to influence not just the Arctic climate, but the climate at other latitudes as well. Thus, a goal of future research should be to assess the relative importance of Arctic change compared to these other regions in determining future mid-latitude weather and its extremes.
The Role of the Stratosphere and Seasonal Snow Cover Paul J. Kushner, University of Toronto
This talk discussed two linkages that figure prominently in discussions of linkages between Arctic and the mid-latitudes: 1) coupling between the polar stratosphere and the extratropical troposphere, and 2) the related coupling between large-scale snow cover anomalies and circulation. Although stratospheric variability is mainly controlled by planetary waves propagating upward from the troposphere, the last 15 years have seen a large number of papers dealing with “stratospheric influence”, which is a back effect of stratospheric variations on the tropospheric circulation. This talk identified two dynamical pathways for this influence on intraseasonal timescales, one involving wave mean flow interactions and the Northern Annular Mode/Arctic Oscillation (NAM/AO), as initially proposed by Baldwin and Dunkerton; and another involving wave reflection and the generation of regional circulation anomalies, as described by Perlwitz, Harnik, and Shaw. In these linkages, the role of the stratosphere is not always direct. In particular, stratospheric
anomalies may have robust tropospheric precursors from sea ice, snow, tropical forcing, etc.. An example of this is the effect of tropical forcing on the Arctic via the polar stratosphere, which is particularly sensitive to the structure of the background stationary wave field through a linear interference effect (papers by Fletcher and Kushner).
Another example of tropospherically forced signals that involve the stratosphere as an intermediary is the proposed role of October Eurasian snow cover in Eurasia in driving wintertime NAM/AO anomalies, as described by Cohen and collaborators. The snow-NAM connection has been found to operate in interannual variability and provide a useful tool for seasonal forecasting. For long-term trends, connections between September sea ice loss, October Eurasian snow increase, and NAM anomalies have been drawn. The talk discussed the role of observational uncertainty and decadal variability in making it challenging to obtain a robust understanding of snow-NAM linkages.
3-Minute Open Session Presentations: Big Picture
Why has the Arctic Warmed? Martin Hoerling, NOAA
A statement from Francis and Vavrus (2012) was presented: “The differential warming of the Arctic relative to mid- latitudes is the key linking Arctic amplification with patterns favoring persistent weather conditions in mid-latitudes.” This was clarified to column warming. Conjectures from the same paper were also provided: observed integrated 1000-500mb Arctic warming is due to Arctic sea ice loss,” and “observed weaker poleward thickness gradients are due to Arctic sea ice loss.” A model was run both with and without sea ice to find the net effect of the ice. The estimated observed 1000-500hPa Arctic warming found was inconsistent with model internal atmospheric variability alone, was largely forced, but inconsistent with effects of sea ice loss alone, and thus could be largely due to natural decadal sea surface temperature (SST) forcing.
Arctic Contribution to Future Storm Track Uncertainty Tim Woollings, University of Oxford
Two strands of relevant work were summarized. The first of these is a set of experiments with one atmospheric model (HadGAM2) to investigate sources of spread in CMIP model projections of the Atlantic storm track response to climate change. This work focuses on the late 21st century and shows that uncertainty in the magnitude of sea ice retreat is a large source of uncertainty in predicting the response of the storm track. The second strand of work looks at future changes in temperature variability. Horizontal temperature gradients are projected to weaken in the future, partly due to Arctic Amplification, and this may result in a reduction in temperature variability in the mid-latitudes in winter, since the effect of wind blowing across temperature gradients will lead to weaker temperature anomalies.
Open Panel Synopsis John Walsh, University of Alaska, Fairbanks
This Open Panel presentation included two items. The first was a summary of the upcoming National Climate Assessment’s (2013) message on linkages between Arctic sea ice loss and mid-latitude circulation. This message highlights the Francis and Vavrus (2012) finding that decreases in subpolar westerlies are correlated with Arctic sea ice loss, but cautions that such conclusions are dependent on the metric used for wave activity and are limited by the small sample size. The second item was a figure showing that 500 hPa zonal wind speeds
averaged over 30-80ºN for the months of October-December have indeed weakened from 1979-present, consistent with summer sea ice loss. However, an extension of the time series back to 1948 shows an increase of zonal winds from the 1950s through the late 1970s. There is no known corresponding trend of (increasing) sea ice cover during the summer/autumn period in these decades.
The Other High Latitude Boundary Forcing: Snow Judah Cohen, Atmospheric and Environmental Research
Though the dramatic reduction in sea ice and its impact on mid latitude weather especially extreme events has received much of the attention, Siberian snow cover has also been shown to influence mid-latitude winter weather. Comparison of correlations between October Siberian snow cover (as measured by the snow advanced index or SAI) and September Arctic sea ice extent and the winter Arctic Oscillation, the dominant mode of variability for the extratopical Northern Hemisphere, shows that the SAI is much more highly correlated with the winter AO than sea ice. An operational statistical model, which uses the SAI as its main predictor, correctly predicted the cold winter in 2013 across Northern Eurasia and the United States (pattern correlation between predicted and observed temperatures of 0.65). Furthermore the statistical model greatly outperformed a suite of dynamical models, which all predicted warm temperatures. The excellent forecast is consistent with other recent accurate forecasts.
Importance of Arctic Linkages for Policy and Communication Benjamin J. DeAngelo, U.S. Environmental Protection Agency (EPA)
This submission offers a client perspective. There are two areas where EPA is engaged where this is relevant.
The first involves communication. EPA produces its periodic “Climate Change Indicators in the United States” (http://www.epa.gov/climatechange/science/indicators/index.html), which includes loss of Arctic sea ice as one of 26 indicators. Loss of Arctic sea ice already serves as an iconic communication tool; it can indicate the rate of climate change as well as ‘how we’re doing’ in terms of GHG trajectories. Linkages between Arctic sea-ice loss and changes in weather patterns across the U.S. are an additional element that could be added to the Arctic storyline for communication purposes.
The second involves EPA’s work under the Arctic Council—a forum of eight nations holding Arctic territory—to address short-lived climate forcers, primarily black carbon. The Arctic Council is working towards some sort of “arrangement” on short-lived climate forcers. Here too, though this NRC effort is not focused on causes of Arctic sea-ice loss (where black carbon has been identified as a possible contributor), further information about the implications of sea-ice loss could inform Arctic Council efforts to address climate pollutants.
SESSION #2 OBSERVATIONAL EVIDENCE OF LINKAGES
Arctic Connections to Extreme Weather: Evidence and Gaps Jennifer Francis, Rutgers University
This presentation discusses evidence, and lack thereof, supporting the proposed chain of events that connects observed rapid warming of the Arctic with changing weather patterns in the mid-latitude northern hemisphere. Some links in this chain are well supported: 1) Arctic amplification (i.e., enhanced warming of the Arctic relative to the northern
hemisphere as a whole), 2) a weakened poleward temperature (or atmospheric thickness) gradient as a result of the differential warming, 3) weakened upper-level zonal (west-to-east) winds in areas where poleward gradients have weakened, and 4) more meridional (i.e., (larger fraction in the north-south direction) upper-level winds in the same areas. The remaining three links in the chain are less solid, but the available evidence is generally supportive. More research, using both new analysis techniques of observations/model output as well as targeted modeling experiments, is clearly needed to either confirm or disprove the links. These include: 5) a more meridional upper-level flow leads to increased amplitude of large-scale (Rossby) waves and an increased likelihood of blocking, 6) a more amplified flow causes west-to-east wave propagation to slow, and 7) slower moving upper-level waves cause more persistent weather patterns, which increase the likelihood of extreme weather events associated with prolonged weather conditions. Additional research is also needed to untangle interactions among Arctic amplification and other large-scale circulation features, such as ENSO/PDO, AO/NAM/NAO, AMO, stratospheric circulation changes, etc.
The Link between Tropical Convection and the Arctic Warming on Intraseasonal
and Interdecadal Time Scales
Steve Feldstein, Pennsylvania State University
During the past several decades, the largest signal of global warming in the atmosphere has been observed to occur over the Arctic Ocean. Coincident with this warming has been a large downward trend in Arctic sea ice. Several theories have been presented to explain this so-called Arctic Amplification. Here, evidenced is presented that the Tropically-Excited Arctic warMing (TEAM) mechanism makes an important contribution to this warming of the Arctic. The TEAM mechanism proposes that the trend toward stronger and more localized convection over the tropical Indo-Pacific Warm Pool excites poleward propagating Rossby wave trains which warm the Arctic by transporting heat and moisture poleward, inducing sinking motion/adiabatic warming over the Arctic, and increasing the downward flux of infra-red radiation.
Results are presented which show that Arctic warming is associated with particular atmospheric teleconnection patterns that fluctuate on a time scale of 5-10 days (the Pacific/North American and circumglobal teleconnection patterns). A link between the observed multi-decadal Arctic warming trend and these rapidly-evolving intra-seasonal time scale teleconnection patterns is shown to arise through an increase in the frequency of occurrence of these patterns over the past several decades.
An analysis of the impact of Arctic sea ice is also presented. It is first shown that Arctic sea-ice anomalies are associated with the dominant atmospheric teleconnections in the Northern Hemisphere. Furthermore, these sea-ice area anomalies are found to precede the teleconnection patterns with lead times of up to 12 months. This link between Arctic sea ice and the circulation over much of the Northern Hemisphere is found to occur through changes in the strength of the stratospheric polar vortex.
Observational Evidence of Arctic Weather Linkages James E. Overland, NOAA/Pacific Marine Environmental Laboratory
Loss of Arctic sea ice, record negative values of the winter Arctic Oscillation atmospheric circulation index, earlier summer snow melt, and increasing extreme weather events at mid- latitudes—both heat waves and cold snowstorms—have been observed over the last decade. While the length of time series is too short to robustly differentiate Arctic forcing of mid-latitude extremes from random events, it is important to identify examples that support proposed linkages. Here, we focus on early winter and find three regionally different mechanisms operating in eastern North America, northern Europe, and far eastern Asia.
Similar penetrations of cold air into the eastern United States in December 2009, 2010, and 2012 relate to a shift in the long-wave upper-level atmospheric wind pattern. This shift coincides with warmer temperatures and greater geopotential thickness over northeastern Canada, major sea ice loss during October in Baffin Bay, a positive Greenland Blocking Index (greater 500 hPa geopotential heights), and record negative values of the Arctic Oscillation index. Such a combination of events amplified and shifted the climatological atmospheric wind pattern westward and allowed deeper southward penetration of cold air into the US. Northward air flow over Davis Strait acted as a positive feedback to maintain the higher air temperature anomalies. Anomalously cold early-winter weather in northern Europe has a direct teleconnection to loss of sea ice in the Barents and Kara Seas. Upper-level atmospheric circulation in north-central Asia in winter responds to the recent large scale reduction the north-south temperature gradients, as noted by Francis and Vavrus (2012), reducing jet-stream zonal velocities and the penetration of storms into northern Asia from the west, increasing the strength and persistence of the Siberian High, and thus increasing the intra-seasonal probability of multiple cold air events over eastern Asia. Midlatitude attribution remains difficult and controversial as there are complex interactions of Arctic forcing with chaotic mid-latitude flow. One would not expect events to happen the same way every year even with similar Arctic forcing.
Linkages between Extratropical Weather and Arctic Weather John R. Gyakum, McGill University
A. Apparent impacts of extratropical cyclogenesis on an extreme Arctic wind case (September 1999)
We study a case of an extreme storm surge, associated with a mesoscale surface cyclone, which occurred along the Beaufort Sea coastline during September 1999. Precursor meteorological conditions included an extreme water vapor transport into the Alaskan coast, and an extreme case of explosive cyclogenesis event affecting the southern Alaskan coast. A mesoscale modeling study has shown that latent heating associated with the explosive cyclone to the south was responsible for building a meridionally-oriented upper-tropospheric ridge that provides support for upshear surface cyclogenesis in the Beaufort Sea, whose flow contributes to the strong surface winds at Tuktoyaktuk. The local effects of the opening of the Beaufort Sea, though enhancing vulnerability, do not appear meteorologically important for such a storm surge event. If one component of future climate change is an eastward displacement of storm tracks into the northeastern Pacific, we may expect more such episodic events.
B. Apparent impacts of Arctic air mass formation on explosive extratropical cyclogenesis (January-February 1979)
The genesis of an arctic air mass in late January 1979 preceded a record cold-air outbreak in the United States. The associated planetary-scale flow facilitated a succession of synoptic-scale vorticity maxima that were responsible for a week-long clustering of explosive cyclogenesis. During this explosive cyclogenesis clustering, a strong increase in the North Atlantic basin’s coverage of strong moist baroclinic growth rates occurred. Additionally, the largest Northern Hemispheric available energy peak was observed during the period just prior to the onset of the clusters of explosive cyclogenesis. This work shows that the process of arctic air mass generation in northern Alaska has significant planetary-scale consequences.
3-Minute Open Session Presentations: Observations
Open Panel Talk
Xubin Zeng, University of Arizona
Monthly diurnal temperature range (DTR) is simply the monthly average of daily (maximum – minimum) land surface air temperature difference. First, we find the zonally-averaged DTR in January over high latitudes to remain large, which is opposite to our physical understanding as diurnal solar forcing is zero. Further analysis of hourly temperature data indicates that these DTR values represent the synoptic movement of weather systems every few days (rather than the diurnal cycle). A more robust and appropriate metric (DTRh), computed as the amplitude of monthly-averaged hourly temperature diurnal cycle, is proposed to represent the diurnal cycle. Second, we find a negative DTR trend in January from 1979-2009 north of 40oN, consistent with previous studies. However, this does not represent the change of diurnal cycle with time. In fact, DTRh does not show any trend. These results suggest the revision of all previous studies on DTR over high latitudes in winter.
Differential Pole-Eq. Warming and Rainfall Rit Carbone, National Center for Atmospheric Research
Slower tropospheric winds suggest weaker vertical shear of horizontal wind. Vertical shear of greater than or equal to 10-3 s -1 is a necessary condition for 50-60 percent of June-July-August (JJA) rainfall in the central United States. This phenomenon is common to many “breadbaskets” of the world. The NCAR Climate System Model (CSM) 4th Assessment runs from 2090s – 2000s present an “ill-posed finding” on vertical wind shear reduction in that because global climate system models cannot simulate the described phenomena, they likely have substantial systematic errors in winds, momentum fluxes and vertical shear. Therefore the significance of projected decreases in shear, while physically plausible and expected, is uncertain. This finding implies a major redistribution of continental warm season rainfall, both spatially and temporally within the diurnal cycle. This finding may be similarly applicable to major portions of China, South America, Australia, Europe, Africa and South/Southeast Asia. Convection-permitting global climate models are needed to resolve this issue.
Flash Floods and Drought: 1) An Indicator of Shifts in Downpours (Precipitation IDF Curves approach) and 2) 2012 Summer Anomalies and the U.S. Drought Sam Higuchi, NASA
There are two approaches to examining the relationship between an increase in flash floods (Intensity-Duration-Frequency Curves) and Arctic amplification, as a leading indicator of change and as a lagging indicator of change. NOAA’s traditional approach to lagging indicators of change assumes the stationary case, using traditional statistical approaches (based on the central limits theorem) results. At least one site-specific conclusion is based on computational results without looking at the graphed observational data. The nontraditional approach to leading indicators of change is splitting the observational data and graphing observational differences using Generalized Extreme Value theory. The statistical approach suggests non-stationary in some cases (there is a change in slope, similar in concept to “hinge-fit” of “new normals”). NOAA also examined the 2012 U.S. Drought and mid-latitude summer weather anomalies and the short-term weather signatures and long-term hints of climate shifts. The analysis was limited to North America, included no jet stream analysis, and no atmospheric blocking analysis. This should be looked at again based on new evidence.
Warm Arctic—Cold Continents, Common among Melting Sea Ice, Rapid Snow Advance and the Negative AO
Judah Cohen, Atmospheric and Environmental Research
Regressions between autumnal sea ice extent, Eurasian snow cover extent the Arctic Oscillation (AO) and Northern Hemisphere temperatures yield the characteristic ‘warm Arctic – cold continents’ pattern. This pattern was observed during winter 2012/13 and is found to be common among years with observed low fall sea ice, rapid fall Siberian snow cover advance and a negative winter AO. Dynamical models, however, fail to capture this pattern, showing instead maximum warming over the Arctic Ocean and widespread winter warming over the adjacent continents. Plotting the daily-standardized polar cap geopotential height from 10-1000 hPa (PCH) from October 2012 through March 2013 shows that episodic warming of the Arctic coincides with many extreme weather events across the mid-latitudes. Regression of September 2012 sea ice extent anomalies onto the PCH from October 2012 through March 2013, further suggests that the dramatic decrease in sea ice contributed to extreme weather events observed during that period.
Pacific Arctic Sea Ice Loss Kathleen Crane, NOAA
Pacific Arctic sea ice loss from advection of heat from the Pacific and Atlantic Gateways is shown. Added ocean heat storage and heat flux from new sea ice free areas affects also the mid-latitudes. More observations are needed to improve boundary layer conditions in models (e.g., the thinning of ice from below using buoys). There are almost no observations in some key areas for heat transport. NOAA is working with Russian Academy of Sciences to improve these observations in the Arctic Ocean.
Probability Density Function of Standardized Anomalous Daily Sea Level Pressure–Based Index of the North Atlantic Oscillation
Matt Newman, University of Colorado
The probability density function (PDF) of standardized anomalous daily Sea-Level-Pressure based index of the North Atlantic Oscillation (NAO) is estimated for two 68-year periods, 1874 to 1942 and 1943 to 2010, both as raw Histograms and as fitted “SGS” PDFs. Results are shown in a figure for each one of the 56 members of the observational 20th century reanalysis ensemble. There are thus 56 curves in each plot for each period. The spread among the curves is a measure of observational uncertainty, whereas the grey swath is a measure of sampling uncertainty. The figure shown indicated that there is no statistically significant change in the PDF of the NAO from the first to the second 68-year period.
A Synoptic-Dynamic Analysis of the Intense Arctic Cyclone of Early August 2012 Lance Bosart, University at Albany/SUNY
Lance Bosart provided a brief overview of his presentation, “A Synoptic-Dynamic Analysis of the Intense Arctic Cyclone of Early August 2012,” from the American Meteorological Society’s 12th Conference on Polar Meteorology and Oceanography. The full abstract of that presentation is provided below:
A surface cyclone formed along an anomalously strong baroclinic zone over north-central Russia on 2-3 August 2012. This cyclone moved northeastward, intensified slowly, crossed the northeastern coast of Russia on 4 August, and strengthened rapidly as it moved poleward over the Arctic Ocean on 5-6 August. By 1200 UTC 6 August, this intense Arctic Ocean cyclone had achieved a minimum sea level pressure of < 965 hPa near 83 N and 170 W. This presentation is motivated by the likelihood that this cyclone was arguably the most intense storm system to impact the Arctic Ocean in the modern data record going
back to the International Geophysical Year in 1957-1958. The purpose of this presentation will be to present the results of a synoptic-dynamic analysis of this intense early August cyclone to help gain a better understanding of why such intense cyclones are so rare over the Arctic Ocean.
Anticyclonic wave breaking in the upper troposphere across Russia in late July and very early August 2012 created an anomalously strong baroclinic zone between 60-80 N. Between 90 E and the Dateline, negative 850 hPa temperature anomalies between -2 and - 4 C were found poleward of 70-75 N over the Arctic Ocean in the 1-5 August time mean. Likewise, positive time-mean 850 hPa temperature anomalies upwards of 8-9 C were situated over eastern Russia near 60 N. In response to this observed temperature anomaly pattern, an anomalously strong 850 hPa temperature gradient of ~10 C (2000 km)-1 between 60-80 N helped to sustain an anomalously strong (20-25 m s-1) 250 hPa jet along the coast of northeastern Russia. A local wind speed maxima along this 250 hPa jet corridor reached 40-50 m s-1 immediately upstream of the surface cyclone.
Because the surface cyclone intensified most rapidly over the relatively ice free Arctic Ocean in the poleward exit region of the aforementioned jet streak, the question arises as to how much sensible and latent heat fluxes from the relatively ice free Arctic Ocean contributed to destabilizing the lower troposphere and augmenting the dynamically driven component of the observed cyclogenesis. Likewise, unusually high observed 1000-500 hPa thickness values between 564-570 dam in the warm sector of the developing cyclone over north-central Russia were indicative of the strength of the cyclone warm sector and the ability of warm-air advection to sustain deep ascent. We will attempt to distinguish the relative importance of dynamical versus thermodynamical forcing to the cyclogenesis process in our presentation.
Declining Spring Snow Cover Extent over High Latitude Northern Hemisphere Lands David A. Robinson, Rutgers University
Annual snow cover extent (SCE) over Northern Hemisphere (NH) lands has averaged lower since the late 1980s than earlier in the satellite era that began in the late 1960s. This is most evident from late winter through spring, and in the past decade has been exceedingly pronounced at high latitudes in May and June. Monthly SCE is calculated at the Rutgers Global Snow Lab from daily SCE maps produced by meteorologists at the National Ice Center. The most recent four Mays have had four of the five lowest NH SCEs on record, with Eurasian (Eur) SCE at a record low in 2013. North American (NA) SCE achieved a record minimum in May 2010, but of late has not been as consistently low as over Eurasia. The past six Junes have seen record minimum SCEs over NH and Eur, with five of these six the lowest over NA. The recent early timing of arctic snowmelt appears to be occurring at an equivalent if not greater pace than the loss of summer Arctic sea ice extent.
THEORETICAL AND MODELING WORK
Atmospheric Energy Transports, Polar Amplification, and Mid-latitude Climate Dargan M. W. Frierson, University of Washington
A variety of general circulation model studies in recent years have shown that changes in high latitudes affect climate even in faraway regions of the globe. Since many of the mechanisms for such connections are based on energy transports, we review expectations for how the poleward transport of energy into the Arctic might change in the future. In simulations of global warming, increased latent energy transport is compensated by reduced dry static energy transport. The total atmospheric transport is actually
anticorrelated with polar amplification across models, and some of the models with the most polar amplification have anomalous energy transport towards the mid-latitudes. We suggest that the response of energy transports to Arctic amplification is first order diffusive, meaning some of the high latitude warming will be spread into mid-latitudes via reduced dry static energy transport.
However, there still can be some local mid-latitude cooling and enhanced extremes in response to rapid Arctic amplification. We propose a new chain of dynamical links that could connect Arctic amplification to mid-latitude weather extremes: decreased temperature gradients and modified high latitude wind profiles should induce a negative phase of prominent oscillation indices such as the NAO, and we suggest that either baroclinic or barotropic mechanisms could cause such a shift. A negative phase NAO index is associated with regions of local cooling, more Greenland blocking, and more temperature extremes. These dynamical links should be tested in model simulations, and regional and seasonal sensitivities should also be evaluated.
Response of the Wintertime Atmospheric Circulation to Current and Projected Arctic Sea
Gudrun Magnusdottir, University of California, Irvine
Motivated by the rapid September decrease in sea-ice concentrations over the years 2007-2012 (inclusive), Yannick Peings and I recently planned and ran some idealized CAM5 experiments using sea-ice forcing from those 6 years (compared to average sea ice conditions over 1979-2000). Our motivation was in part to isolate effects from these few years of rapid decline in sea-ice forcing from other possible forcing mechanisms in the Arctic that would be present in the observational record. Our secondary motivation was to compare results from this experiment (called 2010C) to another experiment where we forced the model with projected sea-ice concentrations corresponding to 2080-2099 (called 2090C).
The surprising result is that the 2010C experiment (or current conditions scenario) forces a remote atmospheric response in late winter that favors cold land surface temperature over mid-latitudes as has been observed in recent years. Anomalous Rossby waves forced by the sea ice anomalies penetrate into the stratosphere in February and weaken the stratospheric polar vortex, resulting in negative anomalies of the Northern Annular Mode (NAM) that propagate downwards during the following weeks, especially over the North Pacific. The seasonality of the response is attributed to the timing of the phasing between the anomalous and climatological waves. When sea ice concentration taken from projections of conditions at the end of the 21st century is prescribed to the model, negative anomalies of the NAM are visible in the troposphere, both in early and late winter. Little impact is found in the stratosphere in this experiment, and this NAM-signature is driven by the large warming of the lower troposphere over the Arctic. A weak but significant increase of the mid-latitude meanders is identified, however, the thermodynamical response extends beyond the Arctic and offsets the dynamical effect at least north of 45 °N. Thus, the stronger sea ice forcing results in no stronger intensity of cold extremes over mid-latitudes.
This latest version of the NCAR AGCM (CAM5) has been improved in terms of representing the seasonal cycle of Arctic clouds (Kay et al 2012, J Clim, 5190-5207). Perhaps more importantly the stratospheric circulation is truer to observations (was too strong in previous versions of the model). We haven’t really addressed the reasons for the remote response now whereas none was found in older versions of the model.
Arctic Sea Ice Predictability and its Long-term Loss and Implications for Ocean Conditions Marika Holland, National Center for Atmospheric Research
Climate models have been very useful tools for exploring linkages between Arctic sea ice loss and climate. In particular, the loss of Arctic sea ice has implications for ocean circulation in the north Atlantic, with resulting effects on poleward ocean heat transport. The magnitude of these simulated changes is strongly related to the magnitude of ice loss (which is nearly linear with the atmospheric CO2 increase). Using Arctic sea ice predictions to forecast future weather has real limitations because our ability to predict the sea ice itself is so limited.
The seasonal prediction of Arctic sea ice is an initial value problem. With almost perfect knowledge of initial conditions, there are inherent limits on predictability. One can examine how ensemble members diverge over time and compare that to the natural variability of the system; when these are indistinguishable, predictability is lost. The model studies examined in this presentation use simulations from CCSM3. Similar initial conditions were used to do an ensemble of predictions with slightly varying initial atmospheric conditions in each run. Results show that the potential prognostic predictability (PPP) decreases during spring and is regained during summer and following winter. There is significant winter predictability due to memory in ocean heat content. The memory of ice edge location is associated with sea surface temperature (SST) predictability. The model studies suggest limited predictability in sea ice area on 1-2 years timescales, particularly in winter. It is currently unclear what spatial coverage, variables, etc. are needed for “adequate” observational network, as well as what model biases impact predictability and what model complexity is required.
The current opportunities for using Arctic sea ice predictions to assess the risk of temperature and precipitation anomalies and extreme weather events over northern continents are subject to inherent limits of the sea ice prediction system. Sources of uncertainty in projections of future climate change include intrinsic climate variability, and the fact that climate models simulate the statistics of climate and not the events of any particular year. North Atlantic circulation changes lead to reduced ocean heat transport and reduced warming. Regarding the impact of longer-term ice-loss changes on mid-latitude weather, these ocean changes need to be considered. Much of the mid-21st century North Atlantic (and high latitude) surface change can be attributed to summer sea ice loss.
3-Minute Open Session Presentations: Modeling
The Role of Global Climate Change in the Extreme Low Summer
Arctic Sea Ice Extent in 2012
Rong Zhang and Thomas R. Knutson, NOAA/GFDL
Comparisons between observations and 19 CMIP5 models reveal that the 2012 summer Arctic sea ice extent (ASIE) anomaly and the rapid decline of summer ASIE in the early 21st century are very rare occurrences in the context of these models and their responses to anthropogenic and natural forcing combined. The observed 2012 record low in summer ASIE is extremely unlikely to have occurred due to internal climate variability alone, according to the models, and has a much greater likelihood of occurrence in the “forced plus internal variability” scenario. The observed September ASIE decline trend for 2001-2012 is much more rapid than in the previous two decades and even lies outside of the 5th-95th percentile range of the multi-model distribution of forced responses, despite the
observed September global mean surface air temperature (SATgm) warming trend for the same period being smaller than in previous decades.
The Influence of Sea Ice Albedo on the Global Hydrological Cycle Aneesh Subramanian, Ian Eisenman, and Simona Bordoni, Scripps Institution of Oceanography, University of California, SanDiego
The idea that Arctic sea ice retreat could influence precipitation outside the polar region has garnered considerable interest in recent years. As sea ice recedes, it alters the climate system due to factors including a change in surface albedo, which has been suggested to affect the hydrological cycle in the mid-latitudes and tropics. Here we examine how sea ice influences the response of the global hydrological cycle to climate change in the broader context of climates ranging from an ice-covered earth to an ice-free earth. We use an idealized general circulation model with annual-mean forcing and the surface albedo set to an ice-covered value where the temperature is below the freezing point and an ice-free value elsewhere. We assess the importance of sea ice albedo by comparing the simulation that includes ice albedo effects with a second simulation that has the same global-mean surface temperature but has the albedo set to the ice-free value everywhere (even where ocean temperatures fall below the freezing point). This work builds on a body of previous theories developed in idealized frameworks that did not include surface albedo changes. We find that ice albedo effects cause substantial changes to the global hydrological cycle in cold climates, but not in warm climates, with the transition occurring at a point relatively near to the present-day climate. Intriguingly, we find that the inclusion of ice albedo effects in cold climates causes an increase in global-mean precipitation despite a decrease in absorbed shortwave radiation and hence less energy available to evaporate water. Large-scale condensation, rather than convection, is the main contributor to the changes in climates close to present day. We explain this result with the finding that ice albedo effects cause a steeper temperature gradient and hence shorter distances for air parcels to travel along isentropes before reaching saturation (a thermodynamic effect) as well as a more vigorous atmospheric circulation (a dynamic effect). These results suggest that ice albedo effects are important for understanding the hydrological cycle in climates colder than today, including many paleoclimate environments.
Arctic Influence on Weather and Climate in Japan Jinro Ukita, Niigata University
At present there remain unclear as to both mechanisms and implications of linkages between Arctic and mid-latitudes changes in weather and climate. In Japan we have established a research project specifically addressing this potentially significant climate question (http://www.nipr.ac.jp/grene/e/index.html). We are currently investigating possible Arctic impacts on NH weather and climate using a JAMSTEC-AGCM with the model configuration of T79L56 and the model top set at 0.08hPa. The numerical experiment consists of two perpetual runs with prescribed SST and sea ice concentration (SIC) conditions from the 1979-1983 and 2005-2009 periods, respectively. In the case where the only difference in the boundary condition is from NH SIC there appears statistically significant cooling in the eastern Siberia and weaker cooling in the eastern Europe and northeastern North America in the temperature difference field. Further analysis reveals that this is due to southward propagation of the Rossby waves aloft originated from an anomalous heat source over the Arctic.
Open Session Talk Uma Bhatt, University of Alaska, Fairbanks
Daily sea ice concentration tendency in five CMIP5 simulations is compared with observations to reveal that most models underestimate this quantity that describes high-frequency ice movements, particularly in the marginal ice zone. Does underestimating this high-frequency ice variability matter to the atmosphere? A set of global climate model simulations were conducted with prescribed sea ice and demonstrate that the atmosphere responds differently when daily ice variations are included. Sea ice differences in September lead to an anomalous high and weaker storm activity over northern Europe. During October, the Arctic ice expands equator-ward faster with daily ice and leads to a local response of near-surface cooling. In DJF, there is a 1.5-hPa positive sea level pressure anomaly over North America, leading to anomalous northerly flow and anomalously cool continental U.S. temperatures. The differences arising from high temporal frequency ice variability, while modest, are relevant for sea ice impacts on mid-latitude weather.
Is the recent sea ice trend a Rapid Ice Loss Event (RILE)? Cecilia Bitz, University of Washington
RILEs are periods of extreme loss of sea ice extent in a five-year or longer period. We assess the probability of RILEs in the CMIP5 models for 1850-2100 (RCP8.5 scenario) in 84 ensemble members from 25 models. RILEs occur on average 3 times in each ensemble member. Most RILES occur during the first half of the 21st century. The probability of a RILE in a given five year period is about 10% at present, increasing to about 20% by 2040. Before 2040, RILES are increasingly common as the sea ice thins because the radiative forcing is also increasing during the 21st century in the RCP8.5 scenario and because the natural variability of the sea ice is also increasing. RILEs become quite rare in the late 21st century because many models are essentially ice-free at that time. About 5 years after a RILE, there is no lasting impact and the sea ice extent returns roughly to a background rate of loss that is similar to the loss rate prior to the RILE. Hence, there is no evidence of tipping point behavior in the CMIP5 models.
Our analysis indicates that it is possible we have experienced a RILE this decade, and possibly now we are on a trajectory back towards the background rate of ice loss. Our analysis indicates that at a given time, it is impossible to distinguish a RILE from a continuously accelerating rate of loss because the difference depends on the future: A RILE necessarily ends with a period of weaker sea ice loss, while a continuously accelerating loss has increasingly greater loss with each year.
Open Session Talk
Clara Deser, NCAR
We investigated the role of oceanic feedbacks in the atmospheric response to GHG-induced Arctic sea ice loss. To do this, we performed atmosphere-only and coupled atmosphere-ocean model integrations with the same late 21st century Arctic sea ice conditions specified as a lower boundary condition. The future sea ice conditions were taken from a fully-coupled model integration forced with the RCP8.5 GHG scenario. In this way, we were able to isolate the role of ocean coupling in the atmospheric response to Arctic sea ice loss. The results show that the primary effect of having an interactive ocean is to deepen the atmospheric temperature response to the mid-troposphere (about 400 hPa) compared to the atmosphere-only run in which the thermal response is confined to the planetary boundary layer (below 850 hPa). This result has implications for the role of high-
latitude ocean feedbacks in Arctic Amplification. The near-surface atmospheric circulation response did not appear to be sensitive to ocean coupling.