Sea Ice News Volume 5 #4 – 2014 Sea Ice Forecast Contest First Predictions

From the Sea Ice Prediction Network

Summary: Thank you to the groups that contributed to the 2014 June report; 28 pan-Arctic contributions (plus one additional regional) is a record for the Sea Ice Outlook! We also welcome several new groups this year.

The median Outlook value for September 2014 sea ice extent is 4.7 million square kilometers with quartiles of 4.2 and 5.1 million square kilometers (See Figure 1 in the full report, below). 

Contributions are based on a range of methods: statistical, numerical models, estimates based on trends, and subjective information. We have a large distribution of Outlook contributions, which is not surprising given the different observed values in 2012 and 2013. The overall range is also large at 3.2 to 6.3 million square kilometers. The median outlook value is up from 4.1 million square kilometers in 2013. These values compare to observed values of 4.3 million square kilometers in 2007, 4.6 million square kilometers in 2011, 3.6 million square kilometers in 2012 and 5.4 million square kilometers in 2013. Only three outlooks this year are above the 2013 observed September extent.

As the season progresses, it will be interesting to follow the relative influences of weather (i.e., warm Arctic winter in 2013/2014 and the wind and weather through the summer) and initial conditions of the sea ice, all in the context of the long-term downward trend of sea ice extent.

This month’s full report includes the comments on modeling outlook, a summary of current conditions, key statements from each Outlook, and links to view or download the full outlook contributions. We appreciate the addition of recent ice thickness data estimated from the European Space Agency CryoSat-2 satellite, the NASA IceBridge airborne campaign, and Office of Naval Research (ONR) Marginal Ice Zone Program buoys.

This June Outlook report was developed by lead authors, Walt Meier (NASA) and Jim Overland (NOAA), with contributions from the rest of the SIPN leadership team, and with a new section this year analyzing the model contributions by François Massonnet, Université Catholique de Louvain.

Full Outlook Report

Overview

Thank you to the groups that contributed to the 2014 season, 28 pan-Arctic contributions (plus one additional regional) is a record for the Sea Ice Outlook (SIO)! We appreciate the diversity of contributions. The median Outlook value for September 2014 sea ice extent is 4.7 million square kilometers with quartiles of 4.2 and 5.1 million square kilometers (See Figure 1). Thus we have a large distribution of Outlooks, which is not surprising given the different observed values from 2012 and 2013. The overall range is also large at 3.2 to 6.3 million square kilometers. The median outlook value is up from 4.1 million square kilometers in 2013. These values compare to observed values of 4.3 million square kilometers in 2007, 4.6 million square kilometers in 2011, 3.6 million square kilometers in 2012, and 5.4 million square kilometers in 2013. Only three outlooks this year are above the 2013 observed September extent. It is indeed an issue to separate the nominal downward trend in sea ice loss from year to year weather influences. Based on previous years’ SIO, weather in 2007 was a factor on the low side while weather in 2013 was a factor in the high side. More sea ice at the end of summer 2013 may be offset somewhat by a warm Arctic winter in 2013/2014. We appreciate the addition of recent thickness data estimated from the European Space Agency (ESA) CryoSat-2 satellite, the NASA IceBridge airborne campaign, and Office of Naval Research (ONR) in situ buoys.

For the June report, our focus is on the pan-Arctic outlook and so there is not a separate regional report; any regional information is integrated into the full report below. We will focus on regional Outlooks in the August report, and will encourage anyone with interest in regional scale to submit their outlooks at that time.

Figure 1. Distribution of individual Pan-Arctic Outlook values (June Report) for September 2014 sea ice extent (values are rounded to the tenths).Figure 1. Distribution of individual Pan-Arctic Outlook values (June Report) for September 2014 sea ice extent (values are rounded to the tenths).

Download High Resolution Version of Figure 1.

Comment On Modeling Contributions To The Sea Ice Outlook

Provided by François Massonnet, Université Catholique de Louvain

Out of 28 predictions, 10 are based on ocean–sea ice (4) or fully coupled models (6). The median predicted September sea ice extent is 4.65 million km2 (mean: 4.71 million km2). The standard deviation of individual predicted extents amounts to 0.73 million km2. All contributors provided an estimate of uncertainty around their prediction. The uncertainty was in all cases estimated using ensemble simulations. An encouraging point lies in the large number of members provided to estimate the impacts of natural variability on the prediction: on average 20 members are used per group, two groups use less than 10 members and three use more than 30. On average, the confidence intervals provided by each group (the bars in Fig. 2) have a width of 1.25 million km2. This could be interpreted as the inherent uncertainty due to the unpredictable effects of the atmosphere between June and September. Modeling groups use initial conditions obtained either from a previous run forced by atmospheric reanalyses, or initial conditions obtained directly from data assimilation of sea ice concentration or thickness.

The standard deviation between the “best estimate” predictions of all groups (the dots in Fig. 2) is 0.73 million km2, thus a +/- 2-sigma confidence interval based on all predictions would have a width of 2.92 million km2. This is larger than the contribution from natural variability (1.25 million km2, see above) and suggests that prediction uncertainty arising from structural differences (different models used, different configurations, resolutions) dominate over the uncertainty associated to natural variability.

Four groups provide maps in addition to the Pan-Arctic prediction (PDF of the four map figures). Opening of the Arctic pack near the east-Siberian coasts in September is a possibility, but the uncertainty in ice concentration in this region is large in all groups that provide this regional information. If the ice pack happens to open, the passage will probably remain tight (~500 km as estimated visually). Opening of the ice pack in the Canadian Arctic Archipelago is less likely according to these models.

Figure 2. Modeling contributions to the June 2014 Sea Ice Outlook.Figure 2. Modeling contributions to the June 2014 Sea Ice Outlook.

Current Conditions

At the beginning of June, Arctic sea ice extent was about 400,000 square kilometers below the 1981-2010 average and slightly below the 2012 and 2013 extent on June 1 (Figure 3). However, the seasonal extent decline was generally slower than 2012 during late May and early June;the rate of decline was near-normal relative to the 1981-2010 average.

Figure 3. Arctic sea ice extent 1 March–1 October for 2012, 2013, 2014 and 1981-2010 average (+/- 2 st. dev. shaded). The faded items in the key are not included on the graph. From the NSIDC Charctic Interactive Sea Ice Graph, http://nsidc.org/arcticseaicenews/charctic-interactive-sea-ice-graph/.Figure 3. Arctic sea ice extent 1 March–1 October for 2012, 2013, 2014 and 1981-2010 average (+/- 2 st. dev. shaded). The faded items in the key are not included on the graph. From the NSIDC Charctic Interactive Sea Ice Graph, http://nsidc.org/arcticseaicenews/charctic-interactive-sea-ice-graph/.

The spatial pattern of loss this spring has been marked by lower than normal extent in the Barents Sea, average to above average in Baffin Bay and slightly lower than normal extent in the Bering Sea (in contrast to recent years that have had a late melt out in the Bering). By early June, a large area of open water had opened up in the Laptev Sea, substantially earlier than normal. However, within the pack, the concentration has remained high over most of the central Arctic, with some lower values developing north of Svalbard.

Figure 4. Arctic sea ice concentration on 10 June 2014. From PolarView, University of Bremen, http://www.iup.uni-bremen.de:8084/amsr2/.Figure 4. Arctic sea ice concentration on 10 June 2014. From PolarView, University of Bremen, http://www.iup.uni-bremen.de:8084/amsr2/.

>Figure 5a. Sea level pressure for March and April. From NCEP/NCAR Reanalysis.Figure 5a. Sea level pressure for March and April. From NCEP/NCAR Reanalysis.

Figure 5b. Sea level pressure for May to early June. From NCEP/NCAR Reanalysis.Figure 5b. Sea level pressure for May to early June. From NCEP/NCAR Reanalysis.

Arctic-average sea ice loss this spring was contributed primarily by regional loss in the seasonal ice zones of the Bering and Barents Seas due to active meteorological seasons (Figure 5). The sub-Arctic meteorology also contributed to May snow cover (see NSIDC June 3 report). The Arctic had a weak Dipole Sea Level Pressure (SLP) with a low on the Eurasian Side and a high pressure region from north of the Bering Strait across northern Canada; the SLP pattern is typical of the long term average (1981-2010). We cannot say that the spring meteorology temperature advection contributed to early melt in the central Arctic. In contrast the winter-spring temperatures over the central Arctic were higher than normal (Figure 6).

Figure 6. Air temperature anomaly at the 925 mb level for January - April. From NCEP/NCAR Reanalysis.Figure 6. Air temperature anomaly at the 925 mb level for January – April. From NCEP/NCAR Reanalysis.

The PIOMAS sea ice volume estimates, based on a model constrained by satellite extent observations, indicate a total volume through May in line with recent years, though not at a record low (Figure 7). In general, lower early season volume suggests an average thinner overall ice cover and thus a greater potential for a lower September extent. However, May values, especially with recent years being so close together, do not provide a clear indication of potential September conditions.

Sea ice thickness estimates from the ESA CryoSat-2 altimeter (Figure 8) and the NASA IceBridge airborne campaign during March and April (Figure 9) indicate thick ice (5+ m) in the region north of Greenland and the Canadian Archipelago, but much thinner ice (~1 m) in the Beaufort and Chukchi seas, as well as the Barents and Kara seas. CryoSat-2 does show a thicker tongue of ice, ~2.5-3 m, extending from the pole to the eastern Siberian coast. This suggests that this region will be slower to melt out than surrounding areas. Airborne electromagnetic induction thickness surveys by the SIZONet project in the Chukchi and Beaufort Sea show a tongue of multiyear ice (for the first time in the past few years) extending down into this region. This suggests delayed ice retreat due to thicker ice in the region. While melt onset at Barrow was as early as it has been observed in the past 15 years with near complete snow melt in April, a cooler May and June with additional snowfall and few meltponds have pushed back complete melt of the ice cover.

Ice mass balance buoys deployed in the Beaufort Sea as part of the Office of Naval Research (ONR) Marginal Ice Zone Program indicate that surface temperatures have reached the melting point, at least intermittently, in the region, with some surface melt beginning in the southern part of the Beaufort, but little or no melt farther north (Figure 10), http://www.apl.washington.edu/project/project.php?id=miz.

Figure 7. Model/observation monthly average total Arctic sea ice volume. From the University of Washington Polar Science Center.Figure 7. Model/observation monthly average total Arctic sea ice volume. From the University of Washington Polar Science Center.

Figure 8. Sea ice thickness from the ESA CryoSat-2 altimeter, provided by NSIDC and Nathan Kurtz at the NASA Goddard Space Flight Center.Figure 8. Sea ice thickness from the ESA CryoSat-2 altimeter, provided by NSIDC and Nathan Kurtz at the NASA Goddard Space Flight Center.

Figure 9. NASA Icebridge - Snow depth and sea ice thickness data from the Quick Look data product. Image via: http://www.nasa.gov/mission_pages/icebridge/science/sea-ice.html#.U5-QI6hhtga.Figure 9. NASA Icebridge – Snow depth and sea ice thickness data from the Quick Look data product. Image via: http://www.nasa.gov/mission_pages/icebridge/science/sea-ice.html#.U5-QI6hhtga.

Figure 10. Surface air temperature (top), ice depth and temperature (middle) and water temperature (bottom) at Buoy Cluster 2 of the ONR Marginal Ice Zone project deployment.Figure 10. Surface air temperature (top), ice depth and temperature (middle) and water temperature (bottom) at Buoy Cluster 2 of the ONR Marginal Ice Zone project deployment.

Key Statements/Executive Summaries From Individual Outlooks

The Outlooks are listed below in order of lowest to highest predicted September extent. Extent and uncertainty (in parentheses) values are provided below in units of millions of square kilometers unless noted otherwise. The individual outlooks can be downloaded as PDFs at the bottom of this webpage.

Rennie (Public), 3.20 (2.50-3.80), Heuristic
Starting with the April PIOMAS volume distribution and the April NSIDC average ice extent the estimated extent loss for each 10 cm thickness of ice loss is calculated. This calculation is then correlated with the reported 5-day average NSIDC ice extent loss. The calculation shows that the extent loss is closely correlated with the initial thickness distribution until the end of July. However the final September average figure is heavily dependent on August weather. This projection indicates a relatively slow decline in extent until mid-July, followed by a rapid decline in the last half of July, leading to a greater overall decline in extent by the end of July than occurred in 2012. In the short term the projection indicates that 2014 will not repeat the rapid decline from June 1 to 16 seen in 2012.

Beckwith, 3.25 (2.75-3.75), Heuristic
The large year-to-year variability in the sea ice extent makes this sort of prediction very dicey. Behavior of the sea ice over the past winter and the spring and the large positive temperature anomalies in the Arctic (as high as 20 degrees C over large regions in the past winter) suggest that an extent near that of the 2012 minimum may occur again if there is large export of sea ice out to the Atlantic Ocean via the Fram Strait.

Kauker et al., (AWI/OASys); 3.72 (3.30-4.14), Modeling
Estimated from ensemble coupled sea ice-ocean model runs based on atmospheric reanalyses fields from 1994-2013.

Andersen, 4.00 (3.80-4.00), Statistical/Heuristic
The estimate is based on the average years since 2007, adjusted to account for the extent peaks this past spring (short periods where the extent increased).

Reynolds (Public), 4.06 (3.49-4.63), Statistical/Heuristic
Because the decline in extent is due to increasing ease with which open water can be revealed by declining volume, a simple method is used to predict September sea ice extent based on May sea ice volume for the Arctic Ocean from the PIOMAS model.

Hamilton, 4.10 (3.10-5.10), Statistical
The 2014 extent is estimated via extrapolation of a Gompertz (asymmetric S) curve fit to the historical data.

Met Office (Peterson et al), 4.10 (3.10-5.10), Modeling
Using the Met Office GloSea5 seasonal forecast systems we have generated a model based mean September sea ice extent outlook of 4.1± 1.0 million square kilometers. This has been generated using start dates between 30 March and 19 April to generate an ensemble of 42 members.

Caveat: The ensemble mean forecast is but one of many realizations of possible September sea ice extent produced by the seasonal forecast system. Whilst the system is devised to accurately account for the range of possible outcomes, as expressed by our ensemble spread and projection uncertainty, there is still a possibility of the actual outcome falling outside this estimate.

Naval Research Lab (Posey et al), 4.20 (3.70-4.70), Modeling
The ACNFS outlook for September ice extent is 4.2 Mkm2 ± 0.5 Mkm2. The skill of the Arctic Cap model run in forward mode for a season is not yet quantified.

Cawley (Public), 4.27 (+/- 1.13), Statistical
This is essentially just a Bayesian non-parametric method used to extrapolate the trends in the observations of September sea ice in previous years, and incorporates no expert knowledge whatsoever. It is a purely statistical projection.

Blanchard-Wrigglesworth et al., 4.39 (3.89-4.89), Modeling
Our 2014 September sea ice extent forecast is 4.39 ± 0.50 million square kilometers. This is based upon a forecast using the CESM1 model initialized with sea ice anomalies obtained from the PIOMAS model for May for the period 1st May-1st October. The quoted error is obtained from the standard deviation of the ensemble distribution in September.

Our 2014 outlook calls for September sea ice to be slightly-lower than the expected linear value for 2014; in other words, lower than 2013 but unlikely (10% chance) to beat the 2012 record. We still expect it to be one of the lowest values on record. Regionally, we expect a more reduced sea ice cover in the East Siberian/Alaskan regions compared to the Atlantic facing region (Svalbard, Franz Josef)

Slater, 4.45, Statistical/Heuristic
My 50-day forecast (http://cires.colorado.edu/~aslater/SEAICE/) issued on June 6th suggests that 2014 will be near the 3rd lowest rank year on record, which is how I came to derive my estimated extent for this long-lead time. The result contains huge uncertainty and should not be considered reliable as this long-lead method has not been verified.

Dekker (Public), 4.60, (4.15-5.05 standard deviation range), Statistical
Arctic sea ice decline has global implications for Northern Hemisphere weather patterns and Arctic eco systems and wild life alike, and thus it is concerning that our global climate models so far appear to underestimate the observed rate of decline based on albedo amplification of sea ice alone.

My projection method for the decline in Arctic sea ice is based on another variable affecting how much heat the Northern Hemisphere absorbs. One that shows particularly strong correlation with Sept sea ice extent, and which I think has been underestimated in models and media articles alike: Northern Hemisphere snow cover in spring.

Zhang and Lindsay, 4.60 (4.00-5.20), Modeling
Our seasonal prediction focuses not only on the total Arctic sea ice extent, but also on sea ice thickness field and ice edge location. We feel that, for all practical and scientific reasons, it is particularly important to improve our ability to predict the ice thickness and the ice edge. Needless to say, this is a difficult goal. However, we hope that our effort would contribute to this goal.

Suckling, 4.67 (3.64-5.76), Statistical
A statistical model, known as Dynamic Climatology, is used to make the prediction for September sea ice extent in 2014. The prediction is initialised with the mean of the observed sea ice extent for September 2009-2013 and an ensemble prediction is created simply by adding all of the observed changes in the sea ice extent record from one September to the next over the historical period 1979-2013.

The ensemble members are then transformed into a probabilistic forecast distribution using the kernel dressing approach, in which the parameters of the kernels are determined based on the statistical skill over a set of hindcasts, under cross-validation.

Using this approach the mean (50th percentile) of the forecast distribution suggests a value for sea ice extent of 4.67 Million square Kilometers, with a 5-95th percentile range (3.64, 5.76) and ‘likely range’ (33-66%) of (4.38, 4.97).

Barthelemy et al, 4.70 (3.50-5.50), Modeling
Our estimate is based on results from ensemble runs with the global ocean-sea ice coupled model NEMO-LIM3. Each member is initialized from a reference run on May 31, 2014, then forced with the NCEP/NCAR atmospheric reanalysis from one year between 2004 to 2013. Our estimate is the ensemble median, and the given range corresponds to the lowest and highest extents in the ensemble.

Kapsch et al, 4.75 (4.13-5.37), Statistical
For the prediction of the September sea-ice extent we use a simple linear regression model that is only based on the atmospheric water vapor in spring (April/May). Thereby we assume that the spring atmospheric conditions, more precisely the greenhouse effect associated with the water vapor in the atmospheric column, are important for the seasonal prediction of the September sea-ice extent.

Wu et al, 4.80 (4.15-5.45), Modeling
The projected Arctic minimum sea ice extent from the NCEP CFSv2 model with revised CFSv2 May ICs using 31-member ensemble forecast is 4.8 million square kilometers with a SD of 0.65 million square kilometers. The maximum and minimum value for the Arctic sea ice extent from the 31-member ensemble prediction is 3.64 and 5.85 million square kilometers, respectively.

Canadian Ice Service, 4.90, Heuristic/Statistical
Three methods are combined. The first is heuristic forecast based on analyst interpretations of spring conditions. The second method uses a optimal filtering based statistical model, and the third estimate is based on regression models relating September sea ice extent to spring atmospheric and oceanic conditions.

Geophysical Fluid Dynamics Laboratory(GFDL)/NOAA (Msadek et al), 4.95 (4.24-5.55), Modeling
Our prediction for the September-averaged Arctic sea ice extent is 4.95 million square kilometers, with an uncertainty range going between 4.24 and 5.55 million square kilometers. Our estimate is based on the GFDL CM2.1 ensemble forecast system in which both the ocean and atmosphere are initialized on June 1 using a coupled data assimilation system. Our prediction is the bias-corrected ensemble mean, and the given range corresponds to the lowest and highest extents in the 10-member ensemble. Our model predicts that September 2014 Arctic sea ice extent will be 1.57 million square kilometers below the 1981 to 2010 observed average extent, but will not reach values as low as those observed in 2007 or 2012.

Stroeve et al (NSIDC), 4.97 (4.24-5.70), Statistical
We use the survival of ice of different ages to statistically predict the 2014 minimum. Summer survivability of ice age fields gives an estimate of 4.97 + 0.73 106 km2 using an average of ice survival rates from the last 5 summers (2009-2013). To estimate a range, we used estimates from the summer with the lowest (2012) and highest (2009) survival rates within the last 5 years, giving an expected range of 3.97 106 km2 to 5.82 106 km2.

Bosse (Public), 5.00 (4.55-5.45), Modeling
The approach with OHC of the year n-1 as the variable with the most important influence on the melting of the year n is a new one AFAIK. The model is plausible: The melting depends on the quantity of heat of the Arctic basin and the volume of the existing sea ice in the actual year.

Yuan et al, 5.20 (4.08-6.32), Statistical
The Markov model is capable to capture co-variability in the ocean-sea ice –atmosphere system, which is likely the predictable part of variances in sea ice. The model focuses on predicting this part of variances. Cross-validation skill, measured by correlation between four-month lead prediction and observation of September ice extent, is 0.77.

Global Weather and Climate Logistics, 5.32, Statistical
Our predictor screening approach predicts slightly more sea ice extent than last year and our anomaly correlation approach predicts slightly less sea ice extent than last year. Together the eight ensemble statistical forecasts predict 5.32 million square km of pan-Arctic sea ice extent for September 2014.

NASA Global Modeling and Assimilation Office (GMAO)/Cullather et al, 5.34 (4.90-5.78), Modeling
A projection of 5.34 ± 0.44 million km2 is made from forecasts initialized from 11 April until 1 May. The forecast has been bias corrected to NSIDC Sea Ice Index values over the previous 15-year period. This projection is made in order to understand the relative skill of the forecasting system and to determine the effects of future improvements to the system.

Centre for Polar Observation and Modelling (CPOM)/Schroeder et al, 5.40 (4.90-5.90), Statistical
We predict the September ice extent 2014 to be similar to last year. The melt-pond area in May is relatively low due to colder air temperatures and slightly thicker ice in the relevant areas of the Arctic in comparison to the last 5 years.

Slater (“Persistence” Outlook), 5.50 [average of three values], Heuristic
Three different types of persistence forecasting at long lead-time were used. The three values were then averaged. The methods contain no skill.

WattsUpWithThat.com, 6.12, Heuristic
A poll was conducted on June 10th, 2014 (http://wp.me/p7y4l-sVu ) of readers of wattsupwiththat.com regarding the Sept. Average Sea Ice extent. 519 votes were cast with a selection of values ranging from 8 million to 3 million square kilometers. Predictions were supplemented by NCEP/NOAA’s new CFSv2 model that showing an above-normal ice extent for Sept, 2014, seen here:
https://wattsupwiththat.files.wordpress.com/2014/06/cfsv2_capture_june10….

Wang, 6.31 (5.84-6.78), Modeling
The projected Arctic sea ice extent from CPC based on NCEP ensemble mean CFSv2 forecast is 6.31×106 km2 with an estimated error of ±0.47×106 km2.

Additional Regional Outlook

Gudmandsen, Nares Strait Region, Heuristic
A heuristic forecast for the timing of break-up for the ice arch in Kennedy Channel is the beginning of July. The forecast is based on early June temperatures (-3° C) and the formation of narrow leads that normally precede break-up. A large floe in Robeson Channel is also blocking the flow of multi-year ice through Nares Strait. The timing of melt/break-up of this floe is hard to predict.

Individual Outlook PDFs

Attachment Size
Attachment Size
Anderson 35.96 KB
Barthelemy et al. 326.93 KB
Beckwith 46.2 KB
Blanchard-Wrigglesworth et al. 592.52 KB
Bosse 77.04 KB
Canadian Ice Service 922.93 KB
Cawley 58.56 KB
CPOM Schroeder et al. 47.32 KB
Dekker 29.07 KB
GFDL NOAA Msadek et al. 198.48 KB
Global Weather Climate Logistics 215.87 KB
Gudmandsen 428.82 KB
Hamilton 267.97 KB
Kapsch et al. 57.57 KB
Kauker et al. 178.13 KB
Peterson et al. 133.29 KB
NASA GMAO Cullather et al. 2.56 MB
NRL Posey et al. 502.48 KB
Rennie 182.08 KB
Reynolds 247.23 KB
Slater 54.25 KB
Slater Persistence 101.53 KB
Stroeve et al. 627.78 KB
Suckling 107.2 KB
Wang 64.79 KB
WUWT 24.59 KB
Wu et al. 156.58 KB
Yuan et al. 55.54 KB
Zhang and Lindsay 246.3 KB

49 thoughts on “Sea Ice News Volume 5 #4 – 2014 Sea Ice Forecast Contest First Predictions

  1. Alec aka Daffy Duck says:
    June 22, 2014 at 8:39 am

    Top graph is the DEVIATION from the median curve (in this case 2010-2013 previous three years but the graph does not notate that). It shows positive ice retention during the summer ). Air temp has not yet gone above 0 C yet and that is an odd thing all by itself.

    Bottom graph is the ice area in Sq Kl

  2. We are right at 0 C today above 80 Deg. I had missed the day spike above freezing but it is still colder than normal this summer in arctic.

  3. Alec aka Daffy Duck says:
    June 22, 2014 at 9:17 am

    Hmm, I meant why noaa predicting 6.7 million sq km?

    —————————————————————————-

    My bad. I would guess that ambient air temp being below normal, ocean currents being cold and the fact that the polar low is increasing in intensity already have quit a bit to do with that prediction.

  4. 2:1 odds that 2014 will higher than 2013. Only reason not giving higher odds is cannot rule out another Arctic cyclone like in 2012. The natural ice nadir was 2007. AMO has turned. PDO has turned. Pause has continued. 2014 ENSO is looking weak. Nuff said.

  5. Hmmm. Last summer was the coldest on record according to the DMC, and this summer is so far staying very cool. That doesn’t guarantee a higher ice extent / thickness so far but I won’t be surprised if we get higher numbers.

  6. Last year WUWT finished third forecasting 4.8m sq km. The Met Office finished last forecasting 3.4m sq km.

    This year the forecast gap between WUWT and Met Office has increased by 0.6m sq km. Which of these teams will be closer this year? Will the crowd’s native intelligence once again beat modeling? I think it will.

  7. Ice mass balance buoys deployed in the Beaufort Sea as part of the Office of Naval Research (ONR) Marginal Ice Zone Program indicate that surface temperatures have reached the melting point, at least intermittently, in the region, with some surface melt beginning in the southern part of the Beaufort, but little or no melt farther north (Figure 10)

    Surface temperature is misleading and not all that important. If I am in the middle of an ice field it doesn’t matter how much heat I dump into the air, the temperature is never going to rise more than a couple of degrees above freezing until the ice melts. If I add more heat, I will increase the melt rate but the temperature won’t rise because that additional melting will moderate the temperature. So in the middle of an ice field, temperature is not a good indication of available heat energy and I’m not sure air temperature even matters much anyway. The vast majority of the melting is from the bottom up, not the top down, and starts long before surface temperatures get anywhere near freezing.

    For example, a few weeks of -30F can keep sea water frozen at a pretty decent depth. Now begin to raise the temperature. At -10F the surface is still well below the melting point but now the ice begins to thin from the bottom up because the surface temperate is not cold enough to keep it frozen at as great a depth. The freeze line moves upward in the ice. By the time the temperature rises above 32F ice ablation from the bottom up by sea water is happening at a fairly decent rate.

    What matters most, in my opinion, is the sea water temperature, wind direction, and wind speed. Warm ocean waters AND persistent winds favorable to blowing large amounts of ice out of the arctic oceans and throw in a couple of cyclones and you have a recipe for great reductions in ice without much increase in air temperature at all. 2007, for example, saw below normal air temperatures for most of the middle of the melt season BUT we saw persistent winds from the west that pushed a huge amount of ice out of the Arctic. With a change in the AMO, I believe we are going to see more sea ice due to colder Atlantic waters. But trying to forecast how much will be there in any given year is crystal ball stuff. One or two storms can change the outcome tremendously.

  8. The weekly SST based on 1970-2000 reveals a strong cooling in the North Atlantic.

    The Barrents and Greenland seas are still quite warm.

  9. WUWT.com seems to have predicted exactly what my guess was. I wish us luck in the contest.

    All that aside, the sea ice at the poles has never been all that important to me. My wife on the other hand wishes there would be far less ice at both poles. I think she would like to see the end of the ice age altogether. Hmmmm. I don’t think she is going to get her wish in this lifetime.

  10. I think we want to keep our icecaps and shelves. They are quite a nice blanket preventing oceanic heat escape. Without them I speculate that our oceans would lose so much heat that the Sun would not be able to replenish what is lost along the conveyor belt.

  11. I learned a lot from the adventurers skiing up there last winter, and the pictures they sent back. They described the ice between Canada and the Pole as especially smashed up. One fellow called it “crazy ice.” The pressure ridges were taller (greater volume,) and there were more leads freezing over (making extra ice.) Temperatures were still quite cold, and as the leads froze over the salt wasn’t only extruded down as brine but upwards as “salt roses.” One fellow sent back pictures. When the winds get strong that salt gets crumbled by blowing snow, and blows along with the snow without melting the snow until temperatures get warmer.

    I have never heard a discussion about the salt content of the blowing snow. It seems to me that there would be more salt at the surface on winters when there are more leads to form the “salt roses.” If the increase in salt was significant, it seems the snow would start melting at a lower temperature, turning available heat into latent heat and perhaps even causing air temperatures to be a bit lower. Just an idea.

  12. Hitting 6.1 for the average extent during september would require a
    melt from now to then that is unprecidented.

    That is, if we look at all past years and measure the melt from this day forward for all the past years,
    we can see that 6.1 would require a very small melt from here on out.

    Obviously possible, but if one bets on history, the observations, then 6.1 is unlikely.

    graphically it looked like this last month

  13. The world renowned Arctic expert, Professor Peter Wadhams of Cambridge University has firmly predicted that the Arctic would become ice free in 2015 or 2016, but no later. No caveats, no ifs, no buts.

    PROFESSOR PETER WADHAMS
    Guardian – 17 September 2012
    This collapse, I predicted would occur in 2015-16 at which time the summer Arctic (August to September) would become ice-free. The final collapse towards that state is now happening and will probably be complete by those dates“.
    ================
    Arctic News – June 27, 2012
    My own view of what will happen is: 1. Summer sea ice disappears, except perhaps for small multiyear remnant north of Greenland and Ellesmere Island, by 2015-16. 2. By 2020 the ice free season lasts at least a month and by 2030 has extended to 3 months…..

  14. Alec aka Daffy Duck says:
    June 22, 2014 at 8:39 am

    Layman here, could someone explain/comment on noaa’s forecast:

    http://www.cpc.ncep.noaa.gov/products/CFSv2/imagesInd3/sieMon.gif

    I’m surprised at:

    2) how many entries seem to be ignoring this altogether.

    1) how little discussion there is on the Web about this except here and at Joe Bastardi’s Weatherbell comments. I went looking last week trying to give a non-WUWT link to a coworker and failed.

    There has been very little discussion about this. My thinking is that the thick ice north of Canada, an area that doesn’t melt until later in the melt season, is not going to completely melt this season. That would provide the above average ice extent around the time of minimum extent. It will also suggests that people planning to sail through the northwest passage may not even get into it.

    That would be an improvement over last year when people got in and couldn’t get out. I don’t know if an ice breaker crunched a path for them or if the boats are still there.

    I hope someone who knows what they’re talking about comes along and explains it all to me too! From Caleb’s comment, he seems to support my thinking.

  15. Robert Bissett says:
    June 22, 2014 at 9:49 am

    If anyone knows the answer, please respond: What is the upside of polar ice caps and continental ice sheets? Thanks in advance.
    =====================================================================
    Well, the wooly mammoths seem to like the… oh… wait.
    .
    .
    .
    I’m kicking myself for not remembering my number this year. I know it was below 6 million square kilometers, 5.5 I think, but I didn’t make a point of noting my number. Dang-it! Won’t make the same mistake next year.
    Individuals who get it right of course can then tell all the rest of us how they knew what the result would be :o)

  16. Remember, once the continental glaciation starts again and the former ocean water gets piled up on the land, higher latitude oceanfront properties will be destroyed, those closer to the equator will quickly loose value as the oceans recede and they become inland property. If you have any oceanfront property you might as well sell it now for a good price.

    But developers will make out like bandits, as they can build lots of trendy pricey seaside mansions where it’s warmer, and a few decades later they can do it again at the new water’s edge. That’ll be some action once it gets going, better get ready to invest now.

    In fact, one of the hottest building markets will be in the Gulf Valley of Mexico, spanning Texas beef to Florida citrus with New Orleans jambalaya in the middle. And because I’m such a nice person, I’ll help you get in early with a sweet deal on a new home you can happily spend the rest of your years in. I know a man in the business.

    Just send an email with the Subject “Mexico Retirement Funding” to mann@psu.edu and ask how he wants to be sent your climate change investment money. He’ll likely respond quickly.

  17. I am wondering whether WUWT got bamboozled by that NOAA report, and tricked into giving a high estimate, when we see no such high estimate coming from NOAA in the final set of estimates?

    I agree with Mosher that WUWT’s estimate would require a surprisingly slow melt season. I voted for 5.25 myself and was dismayed to see the WUWT consensus.

    Rich.

  18. Well, you can all run around smugly smiling about your 6 million square miles – until you see me floating in my kayak this September at the North Pole – the real north pole, not the magnetic north pole. I may not even bring a wet suit it will be so warm up there.

  19. Mods – was it Anthony’s intention that the entirety of this article post to the main page?

    [Fixed, thanks. ~mod.]

  20. Steve from Rockwood says:
    June 22, 2014 at 3:43 pm

    @Jimbo. “probably”, “except perhaps”…

    Naaaah. Look at the context around those words and phrases. He has dug himself into a hole. I too looked carefully for the caveats but could not find them.

  21. Robert Bissett says:
    June 22, 2014 at 9:49 am says What?

    This thread isn’t the place for such an exchange of views. Why not do a little reading and write up what you find. Maybe it could have a shot at being posted here. Then you would get to read what others think of your findings. I’m looking forward to them.

  22. As of yesterday, the sea ice extent for September according to the NSIDC methodology is tracking for 5.65M km2 (based on the average climatology for the remainder of the year) versus 5.247M km2 last year and 3.63M km2 in the record low 2012.

  23. Ric Werme says:
    June 22, 2014 at 2:26 pm
    My thinking is that the thick ice north of Canada, an area that doesn’t melt until later in the melt season, is not going to completely melt this season. That would provide the above average ice extent around the time of minimum extent. It will also suggests that people planning to sail through the northwest passage may not even get into it.”

    You’ve got my vote Ric. Much of the >2m ice extent will survive. The Northwest Passage, only an idiot would bet it will be navigable this year. There are stretches of 4m ice clogging it that just won’t melt – 100% certainty. Lake Superior ice only recently this June finally melted and cottage country lakes in Ontario and Quebec are super cold – they need warm May, June and July to make any other than the intrepid able to enjoy a swim.

    You are right about the silence on the web concerning NOAA’s expected positive ice extent anomaly. http://www.cpc.ncep.noaa.gov/products/CFSv2/imagesInd3/sieMon.gif

    What do the non scientific CAGW proponents do when they see something like this? They pray it doesn’t come true or put their fingers in their ears and shout Lalalalalaleeloo. I myself was already considering something in the range of NOAA’s predictions by measuring the extent of the 2m+ice.
    The colder than climatology temps north of 80 also support this and maybe more.

  24. phlogiston says:
    June 22, 2014 at 5:34 pm

    richcar1225 says:
    June 22, 2014 at 12:33 pm
    “The weekly SST based on 1970-2000 reveals a strong cooling in the North Atlantic.”

    A look at equatorial east Pacific in this map shows a large area of cold water just south of the equator that could kill off the the advent of El Nino, too. If it fizzles, we may have some cooling of both hemispheres and a cycle of Arctic sea ice recovery that will increase the workload on psychiatrists that have been studying the wrong people in the climate battle. They had it easy with only 3% of the population “sick”, but 97%, that’s a lot of kleenex!!!

  25. MrX said June 22, 2014 at 5:15 pm:

    None of the images show up for me.

    To me, it looked like a wonderful IT fail.

    As I noted, an “Image Hotlinking Disabled” graphic replaced the pictures. Likely too much server traffic for just the images triggered the bandwidth-saving failsafe.

    But the graphic was then called up as much as all those images together, thus a huge surge of traffic for one image. And now there are no images coming from ARCUS. I think the server crashed.

    If you want to throttle back bandwidth by disabling hotlinking of images, why reroute the hotlink to a 628×366 image?

  26. “I am wondering whether WUWT got bamboozled by that NOAA report, and tricked into giving a high estimate,”

    I pointed out several times that the NOAA report was not from a validated model.

  27. “Then we have this other chap following closely ahead and behind Wadhams. He is the acclaimed climate scientist Arctic ice modeller Professor Wieslaw Maslowski.”

    1. Both of these guys are extremist.
    2. the consensus is for a much later date.

    Citing them without citing the consensus would be like taking the two most wacked out votes from the WUWT survey and claiming that those outliers stood for WUWT.

  28. ” and a few decades later they can do it again at the new water’s edge. ”

    It will be much slower than that. In fact, it has already started. Sea levels are about 2 meters lower than they were about 6000 years ago and we already have a lot more ice on land. The glaciated valleys in the Alps were completely ice free 6000 years ago. It isn’t even and steady though. We get warm periods and we get cool periods and from here on out there will be someone claiming that we are causing whatever it is doing at the moment so they can capitalize on it. 20th century warming was continued recover from the Little Ice Age and we probably have as much again to go — if we get there before the next cool period sets in.

  29. Steven Mosher says:

    “Then we have this other chap following closely ahead and behind Wadhams. He is the acclaimed climate scientist Arctic ice modeller Professor Wieslaw Maslowski.”

    1. Both of these guys are extremist.
    2. the consensus is for a much later date.

    Citing them without citing the consensus would be like taking the two most wacked out votes from the WUWT survey and claiming that those outliers stood for WUWT.

    Really? Have the two most whacked out votes from the WUWT survey been parroted by media outlets worldwide as science news? Have the two most whacked out voters from the WUWT survey been quoted and their predictions endorsed by Algore? When the media and Algore were promoting the W&W predictions, did they cite “the consensus”? Did the “consensus’ raise vociferous objection to the publicised views of W&W, as they do when an interview with Anthony Watt’s accidently gets included in a MSM program?

    The sea ice prognostications of the W twins were condoned by the “consensus”, so consensus they are.

  30. Steven Mosher says:
    June 22, 2014 at 9:27 pm
    “I am wondering whether WUWT got bamboozled by that NOAA report, and tricked into giving a high estimate,”
    I pointed out several times that the NOAA report was not from a validated model.
    ————————————–

    Mosher comments like there is a “validated” Arctic sea ice model. The only “validated” ones are in someone’s imagination. The NOAA model actually has more real validation behind it than any of the death spiral models.

  31. “Mosher comments like there is a “validated” Arctic sea ice model. The only “validated” ones are in someone’s imagination. The NOAA model actually has more real validation behind it than any of the death spiral models.”

    you obviously havent read the documentation on the model

  32. “Really? Have the two most whacked out votes from the WUWT survey been parroted by media outlets worldwide as science news? Have the two most whacked out voters from the WUWT survey been quoted and their predictions endorsed by Algore? When the media and Algore were promoting the W&W predictions, did they cite “the consensus”? Did the “consensus’ raise vociferous objection to the publicised views of W&W, as they do when an interview with Anthony Watt’s accidently gets included in a MSM program?

    As a matter of fact yes the media did cite the consensus.

    example

    http://knowledge.allianz.com/environment/climate_change/?2432/New-climate-change-report-on-thin-ice

    the way the story was framed was the consensus is TOO conservative.

    Second, I would agree that the media and Al gore spend too much time highlighting the extremes.
    Thank you for making my point. You see these are two side of the same coin: highlighting extremes –whether to alarm or discredit– is the same shoddy misleading unscientific BS.

    the MSM highlighted the extreme to alarm
    folks here highlight the extreme to discredit

    Thank you for making my comparison more forceful! you pretty much own goaled yourself there

  33. Looking at the sea ice page for extent data, that puts my estimation in the range 4.8-5.8, with my most likely estimation being to the top of that range, say 5.5-5.6.

    Now that seems to concur with a lot of the professional estimations, except that they mostly run a lot lower, ie they’re giving a notably wider range of possibilities.

    Anyway there does not seem to be much of the “death spiral” talk apart from crazies like Prof. Wadhams who just wants us to let him play being an apprentice deity with our climate.

  34. “How would you classify PIOMAS? Wacked out? Or validated?”

    More like invalidated. Like most modelling ventures it’s the preconceived ideas of modellers “tuned” into the form of a model.

  35. Steven Mosher says:

    As a matter of fact yes the media did cite the consensus.

    example
    http://knowledge.allianz.com/environment/climate_change/?2432/New-climate-change-report-on-thin-ice

    Please. Some obscure blog that no one has heard of is not “the media”. The major news outlets have squawked the death spiral stories from those asshats. Where was the identification of those predictions as being outside of the “consensus”, by either the author or by response of the “consensus”? Non-existent. If it is condoned by the consensus, it is owned by the consensus.

    Second, I would agree that the media and Al gore spend too much time highlighting the extremes. Thank you for making my point. You see these are two side of the same coin: highlighting extremes –whether to alarm or discredit– is the same shoddy misleading, unscientific BS.

    No. Not even close. When one side cites an extreme view in multiple major global press releases as if it were consensus and/or a likely event instead of the alarmist scare story that it is, then that is shoddy, intentionally misleading, unscientific political propaganda. When the “Climate Rapid Response Team” doesn’t rapidly respond to such extreme views that favor their political goals, then that is shoddy, intentionally misleading, unscientific political propaganda.

    On the other hand, when the other side forces them to own their propaganda by quoting their shit back to them with the observational refutation attached, that is the correction to shoddy, intentionally misleading, unscientific political propaganda.

  36. Steven Mosher says:
    June 22, 2014 at 9:36 pm
    “Then we have this other chap following closely ahead and behind Wadhams. He is the acclaimed climate scientist Arctic ice modeller Professor Wieslaw Maslowski.”

    1. Both of these guys are extremist.
    2. the consensus is for a much later date.

    Citing them without citing the consensus would be like taking the two most wacked out votes from the WUWT survey and claiming that those outliers stood for WUWT.

    Interestingly both them work with the submarine records, Wadhams with the RN data and Maslowski with the USN.

  37. Not looking good, folks. Looks like the Arctic ice is heading for a crash that will easily beat 2012.

Comments are closed.