Too late now for 2013 to produce Greenland ice surface melting anywhere near as large as in 2012

July 22nd, 2013

We’ve made it to the middle of the melt season 2013. It’s been a roller coaster for Greenland climate in 2013.

The year started out with an astonishing low albedo from a snow drought that made huge melt possible. Yet this was punctuated by the return of snow late April and prolonged low temperatures as the North Atlantic Oscillation (NAO) shifted from persistent negative that had been causing west Greenland temperatures to be abnormally warm while Copenhagen thad its coldest late winter and early spring in decades. The dipole in temperatures between NW Europe and W Greenland has been recognized for more than a century and called the “seesaw” in temperature (van Loon and Rogers, 1978).

Like in 2012, melt came on strong by early June but was shut down again by abnormally low temperatures, accompanied by additional snowfalls [1, Ruth Mottram, Danish Meteorological Institute, personal communication] in some areas, that lasted until mid July. The key difference between 2013 and the previous 6 summers (2007-2012) is the absence of a persistent negative NAO that drove south air over Greenland, heating it while promoting clear skies that maximized the impact of surface darkening through the albedo feedback (Box et al. 2012). With this much of a delayed start, the albedo feedback has not had enough time to produce strong melt. Given now that we are at the mid point of the melt season, it is too late now for 2013 to produce melting anywhere nearly as large as we saw in 2012.

The NASA MODIS data indicate that the 2013 Greenland mid melt season (mid-July) albedo is at its lowest in 4 years; behind 2009, 2010, 2011, and 2012.

The 2013 June ice sheet is more reflective across the southern third than the average of the recent decade. This may also be partly a consequence of the late season snowfall which was concentrated in the south eastern quarter (Ruth Mottram, DMI). Across the northern 2/3 of the ice sheet, the difference is below average, though not as far below average as in previous years, especially 2010, 2012.

But the bigger story seems to be the preliminary July 2013 average (the first 20 days of the month) where much of the lower elevation ablation area is more reflective than the recent decade (2000-2011). This is slowing down accumulated melt. Meanwhile, the upper elevations (inland) have below average reflectivity. Why are the upper elevations (inland) somewhat below normal reflectivity? While Dark Snow project was successful in gathering surface reflectivity samples from southwest Greenland this 8th and 9th July, the samples were not from where the high July albedo anomaly is. Were it anyway North American fire smoke contributing to the reduced ice reflectivity, it could be a.) lower than normal snowfall at the upper elevations over the year . Could the albedo of 2013 literally be reflecting the ice of 2012 buried less than 1 m below the surface? The albedo signal can originate from some 10s of cm below the surface. According to DMI’s HIRLAM model estimates, there is less than 1 m of snow accumulation over much of the higher elevations. Then again, the fire factor may also be at play. We’ll be looking into this with Dark Snow project.

  Works Cited

  • [1] Ruth Mottram, Danish Meteorological Institute (DMI), personal communication. The DMI High Resolution Limited Area Model (HIRLAM) simulates snowfall in southeast and east Greenland. The cumulative surface mass balance indicated net mass accumulation. HIRLAM is initialized by observations from satellites, weather balloons, aircraft, ground stations, buoys, etc.
  • Box, J. E., Fettweis, X., Stroeve, J. C., Tedesco, M., Hall, D. K., and Steffen, K.: Greenland ice sheet albedo feedback: thermodynamics and atmospheric drivers, The Cryosphere, 6, 821-839, doi:10.5194/tc-6-821-2012, 2012. open access
  • van Loon, Harry, Jeffery C. Rogers, 1978: The Seesaw in Winter Temperatures between Greenland and Northern Europe. Part I: General Description. Mon. Wea. Rev., 106, 296–310. doi:<0296:TSIWTB>2.0.CO;2

Quebec fires in NASA CALIPSO data, drifting to Greenland

July 21st, 2013

On the lower limit of this 4 July NASA CALIPSO laser scan, between latitude 42.60 N, longitude -68.93 W, evident is a rising smoke plume. The plume seeds cloud formation toward Greenland, reaching the southwest of the island.

The orange areas are smoke aerosols.

On 12 July, when Dark Snow scientist McKenzie Skiles, after camping, was picked up along Greenland’s longest road for a ride back to town, without prompt the driver remarked on haze in the sky from Canadian fires. McKenzie Skiles “It was the first thing he said after I got in the car, as if apologizing for the haze in the air (which was noticeable).”

hazy Kangerlussuaq, West Greenland. Photo McKenzie Skiles

9 July, 2013, we gathered snow and ice core samples from the surface and down through the 2012 melt layer, and we left only footprints. We’ll eventually see how much soot the laboratory and field spectral reflectance measurements tell us is there.

GRACE-cast: Greenland ice sheet mass loss *this week* turns toward average

July 6th, 2013

In my new position, Professor of Glaciology at the Geological Survey of Denmark and Greenland (GEUS), I’ve started to gather momentum working with a fine group of people to establish for the first time, a Greenland ice sheet total mass balance product that estimates what the GRACE satellite measures and posting the estimate on-line 2-3 months ahead of the GRACE processing.

Our new “Nowcast” of Greenland ice sheet mass balance has as little as 24 h delay from realtime.

The product exploits the fact that on average, 90% of the time, when monthly all ice sheet reflectivity (also called “albedo”) goes up, the rate of total ice sheet mass change goes down and visa versa.

The Nowcast only works in the sunlit period from mid April to mid September.

So, now as we’re mid-melt season (6 July), we finally have some interesting news…delay of the 2013 melt season due to relatively northerly air flow along west Greenland has led to the lowest elevations of the ice sheet having an above average reflectivity due primarily to more persistent snowcover (snow patches) and secondarily due to summer snowfall in some areas. According to the empirical relationship, the rate of ice melt water loss from the ice sheet to the surrounding seas has declined and now approaches the average of the 2003-2009 period.

People involved with what I like to call the “GRACE-cast” include: GEUS glaciology post-doc William Colgan; Danish Meteorological Institute (DMI) ice climatologists Dr. Peter Langer and Dr. Ruth Mottram, and Danish Tecnhical University (DTU) geodesists Dr. Valentina Barletta and Dr. Rene Forsberg.

Technical details are documented in a .pdf file at the new product’s new home:
The cumulative sea level contribution is also updated on the new Danish web site.

Greenland ice sheet ablation area albedo above average, upper elevations below average

July 6th, 2013

A delayed melting at low elevations and probably some summer snowfall blanketing the Greenland ice sheet ablation area with (highly reflective) fresh snow have resulted in an important slow down of Greenland melting. This pattern is in contrast to this time last year (in 2012) when record melting was emerging.

Greenland steep fluctuations of 2013 warm and cold

July 4th, 2013
Greenland melt of 2013 year has had fits and starts.

It’s been dipping in and out of abnormal warmth and cold.

The drama began with very low pre-melt albedo March to -mid-April due to a snow drought that made high melt in 2013 seem more than likely. Then, an about face, a lot of snow and relatively cold weather washed over Greenland for the next 6 weeks (20 April – early June)!

Melt then came on strong 3 June yet was punctuated 22 June by a return of cold weather that has remained in place and is forecast through at least 8 July. 
It now seems more than likely 2013 won’t hit 2012 melt record, this after 6 summers in a row of negative North Atlantic Oscillation that favored Greenland heating. The persistence of that pattern had me wondering if, for example, the drop in Arctic sea ice or the complete ablation of snow cover on land had ~permanently altered large scale atmospheric circulation. Yet, what we see with 2013 suggests a more complex situation with extreme fluctuations of warm and cold.

Greenland – an albedo feedback laboratory

June 30th, 2013

Surface reflectivity of sunlight is called “albedo”. Albedo is a Latin-based word referring to whiteness. The higher the albedo, the more sunlight can be reflected. As albedo decreases, more sunlight can be absorbed.

Snow and ice impurities concentrate in “cryoconite” holes on the Greenland ice sheet surface. Photo. J. Box

The absorption of sunlight is the largest single source of melt energy on the Greenland ice sheet.

Surface albedo across Greenland is mapped using data from NASA’s Moderate Resolution Imaging Spectroradiometer (MODIS) satellite-borne sensors. Before melting is underway, albedo is above 80%.

The NASA albedo data have an accuracy better than 5% (Stroeve et al. 2006; Box et al. 2012).

During melting, the rounding of ice crystals by heating causes the albedo to drop.
A freshly fallen snow crystal has numerous facets to reflect sunlight (left). Warming causes the grains to round at the edges and clump together (right). Scanning electron microscope photos courtesy the Electron and Confocal Microscopy Laboratory, USDA Agricultural Research Service.
In some areas of the ice sheet, by the time winter snow cover melt away, bare glacier ice is exposed. Where impurities congregate, the surface albedo drops below 30%.
Aerial oblique view of the lower elevations of the ice sheet in August 2005 from an area near the point of lowest reflectivity on the ice sheet. Photo J. Box

Impurities are composed of dust, algae, wildfire soot. Their relative importance to surface albedo remains incompletely understood.

As part of Dark Snow Project’s 2013 expedition, Dr. Marek Stibal gathers samples from an area of concentration near the darkest point on the Western Greenland ice sheet.

An increase in atmospheric heating of Greenland ice is a driver of Greenland ice albedo decline in summer, in part due to the expansion of bare ice areas, in part due to the heating effect on rounding ice crystals, and in part if the concentration of impurities increases.

In the period of high quality observations beginning early 2000, June 2013 albedo for the ice sheet is ranked 3rd lowest.

Greenland albedo started out very low in 2013 due to a snow drought exposing darker bare ice around the ice sheet periphery.

The albedo feedback with climate is responsible for doubling the temperature changes when climate warms or cools. This amplifier helps Earth’s climate system swing into and out of ice ages. The feedback is complex, including the effects of heating and light absorbing impurities, in a process that compounds through time.

Light absorbing impurities like black carbon from wildfire and industrial sources acts like a multiplier of the albedo feedback.

The Dark Snow Project aims to better understand the black carbon aspect of the albedo feedback through field data gathering and laboratory analysis.


Click here to visit the Dark Snow Youtube Channel.

Works Cited

  1. Box, J. E., X. Fettweis, J.C. Stroeve, M. Tedesco, D.K. Hall, and K. Steffen: Greenland ice sheet albedo feedback: thermodynamics and atmospheric drivers, The Cryosphere, 6, 821-839, doi:10.5194/tc-6-821-2012, 2012. open access
  2. Stroeve, J.C., Box, J.E., Haran, T., 2006: Evaluation of the MODIS (MOD10A1) daily snow albedo product over the Greenland ice sheet, Remote Sensing of Environment, 105(2), 155-171.
  3. Stibal, M. M. Šabacká, and J. Žárský, Biological processes on glacier and ice sheet surfaces, Nature Geoscience 5, 771–774, 2012, doi:10.1038/ngeo1611

Dark Snow Project – stuck in an idyllic place after bagging our first set of samples

June 29th, 2013

As we landed in Greenland 24 June at the beginning of Dark Snow Project, the best laid plans lept out of reach.

Our helicopter was grounded by red tape.

Reacting, several phone calls produced a single flight with a different company. So, Dark Snow Project generated some field data already 25 June and we have another 12 days to work out a way to get back onto the ice sheet to gather more snow and ice samples to document the impact of light absorbing impurities on enhanced ice sheet solar heating.

“I never worked on a meticulously planned ambitious project that didn’t turn into an improvised as-you-go masterpiece.” – Dark Snow Project patron

Dr. Marek Stibal gathers “sediment” from an area of concentration near the darkest point on the Western Greenland ice sheet

Now, awaiting paperwork to push through Danish authorities [Don’t hold your breath. The Dane’s put a work firewall around their weekends], we are using our time productively, collaborating with journalists and scientists.

Our a little blue house, where Dark Snow Project incubates its latest ideas

Peter Sinclair, Sara Penrhyn Jones, and I are gaining momentum editing video. We’ll be filling our You Tube Channel in the coming days.

Follow our Facebook Page and our tweets.

More soon.

end of snow drought and persistent sub-freezing for W Greenland

May 26th, 2013

I’d reported on a highly abnormal snow drought that with more bare ground produced large negative albedo anomalies along west Greenland (Fig. 1).

Figure 1. Greenland reflectivity below 500 m elevation, including land areas. Notice the extreme low anomaly for 2013 that is by now erased.

Well, after about 4 months (1 Jun – 20 April) of that type of anomaly, the pendulum swung back late April, 2013, delivering a ~5 week return of snow showers that brought up to 300% of the normal snow for that period (Fig. 2) and relative cool weather (Fig. 3).  

Figure 2. End of snow drought. Blue and purple areas indicate abnormally high precipitation.

The snow drought is not actually ended everywhere. Along northeast Greenland, snow accumulation remains well below normal, 20% of normal for 1-Jan – 25 May. A @Promice_GL field workers had to transport from Zackenberg station to AP Oleson ice cap using a Argo track vehicle instead of snowmobiles.

Figure 3. Persistent cold for Greenland between 24 April and 19 May.

With the exception of melting 21-25 May, cold has been in place since 24 April. It’s clear now from the forecast for early June 2013 that temperatures will remain below freezing along much of west Greenland. It’s not extremely cold, just not yet melting much.

3 part study reconstructs Greenland ice sheet mass budget since 1840 and presents a theory connecting surface meltwater with ice deformational flow

May 16th, 2013

It took 7 years to pull together a full ice sheet mass budget closure based on a fusion of observationally-based records from coastal and inland weather station temperature readings; ice cores; and regional climate modeling.

Below are links to pre-prints. The papers’ abstracts capture key results but are constrained by 250 word limits. I emphasize part III below. Parts I and II were foundational works with interesting aspects. Part III brings in necessary data from parts I and II.

Greenland ice sheet mass balance reconstruction. Part I: net snow accumulation (1600-2009)
Jason E. Box, Noel Cressie, David H. Bromwich, Ji-Hoon Jung, Michiel van den Broeke, J. H. van Angelen, Richard R. Forster, Clement Miège, Ellen Mosley-Thompson, Bo Vinther, Joseph R. McConnell
Abstract . PDF (3210 KB)

Greenland ice sheet mass balance reconstruction. Part II: surface mass balance (1840-2010)
Jason E. Box
Abstract . PDF (3322 KB)

Greenland ice sheet mass balance reconstruction. Part III: marine ice loss and total mass balance (1840-2010)
Jason E. Box, William Colgan
Abstract . PDF (2452 KB)

Paper III puts forth a theory* linking surface melting with ice flow dynamics. The two are by now too often examined in isolation. Our not so old science of glaciology, beginning in earnest in the late 1950s, can now begin unifying surface and ice dynamics processes at the ice sheet scale. In stark contrast to the messaging that the recent Nick et al modeling study produced, we may expect plenty more sea level contribution from Greenland than current models predict. The misreporting of otherwise good science refers to ice flow to the sea as “melt”. Ice deformational flow is a distinct process from melt. Yet, melt and ice deformational flow are in fact intertwined processes. Self-reinforcing amplifying feedbacks outnumbering damping feedbacks by a large margin (Cuffey and Patterson, 2010, chapter 14) ensure that given a climate warming perturbation, a.k.a. the Hockey Stick, we’ll see a stronger reponse of ice to climate than is currently encoded by models. More on that later.

Because the peak statistical sensitivity between meltwater runoff and ice flow discharge emerges at the decade scale (11-13 years), it seems that the ice softening due to more meltwater in-flow to the ice sheet, the Phillips Effect, if you will, is a central physical process behind a link between runoff and ice flow dynamics (Phillips et al. 2010; 2013).

Yet, on shorter time scales and resulting from a rising trend of surface melting, also to be considered is the effect of meltwater ejection at the underwater front of marine-terminating glaciers. The effect is to force a heat exchange between the glacier front and relatively warm sea water with the ice  (Motyka et al. 2003), melting it. This is, if you will, the Motyka Effect. Underwater melting undercuts the glacier front, promoting ice berg calving and thus providing a direct and immediate link between surface runoff and ice flow. Calving reduces flow resistance, causing ice flow acceleration.

January-February 2013, As I responded to 3 critical anonymous external reviewers and the sands of time were running low to make the 15 March, 2013 deadline for the IPCC AR5, in a bid to increase the likelihood of paper III’s acceptance, I brought on Liam Colgan. His fresh and sharp eyes would comb out any potential text and methodological snags from my major revision. While you may know Liam to be a frequent user of Monte Carlo methods, I already had that in this paper before thinking of his involvement. To his credit, Liam contributed the crevasse-widening aspect to the theory that builds coincidentally via Colgan et al. (2011)’s building on Pfeffer and Bretherton (1987). The Colgan Effect is thus the 3rd aspect of the unified theory this part III study puts forth.

As to the result of the mass budget reconstruction, it’s not surprising that Greenland ice sheet contribution to sea level has accelerated. After all, climate has emerged from the dim-sun Little Ice Age into the greenhouse gas-forced new post-industrial climate epoch, the Anthropocene. Greenland’s going. It’s a question of how fast. I’m happy to report that more to this story is in the works. So, stay tuned.

* A theory is a broad collection of knowledge based on hypotheses (emphasis plural) that have withstood skeptical inquiry and are accepted, unless otherwise proven, as Fact.

J. Climate Editor Anthony J Broccoli deserves thanks for, presumably, working extra in recognition of critical timeline.

Works Cited
  • Colgan, W., K. Steffen, W. McLamb, W. Abdalati, H. Rajaram, R. Motyka, T. Phillips, and R. Anderson, 2011a: An increase in crevasse extent, West Greenland: Hydrologic implications, Geophy. Res. Lett. 38, doi:10.1029/2011GL048491
  • Cuffey, K.M. and W.S.B. Paterson (2010). The Physics of Glaciers, Fourth Edition. Elsevier, 693 pp.
  • Motyka, R. J., L. Hunter, K. A. Echelmeyer, and C. Conner, 2003: Submarine melting at the terminus of a temperate tidewater glacier, LeConte Glacier, Alaska, U.S.A. Ann. Glaciol., 36, 57-65.
  • Nick, F.M., A. Vieli, M.L. Andersen, I. Joughin, A. Payne, T.L. Edwards, F. Pattyn & R.S.W. van de Wal, 2013, Future sea-level rise from Greenland’s main outlet glaciers in a warming climateNature 497, 235–238 (09 May 2013) doi:10.1038/nature12068
  • Pfeffer, W. and C. Bretherton, 1987: The effect of crevasses on the solar heating of a glacier surface, IAHS Publication, 170, 191-205. 
  • Phillips, T., H. Rajaram, and K. Steffen, 2010: Cryohydrologic warming: A potential mechanism for rapid thermal response of ice sheets, Geophys. Res. Lett., 37, L20503, doi:10.1029/2010GL044397.
  • Phillips, T., W. Colgan, H. Rajaram and K. Steffen. Evaluation of cryo-hydrologic warming as an explanation for increased ice velocities near the equilibrium line, Southwest Greenland. J. Geophys. Res. ,2012JF002584, submitted 7 July 2012, revised 31 December 2012.

Greenland “snow drought” makes big 2013 melt more likely

May 3rd, 2013

A friend in Greenland’s capital Nuuk reported (with a frown) that the backcountry skiing this year was poor due to a “snow drought”.


Figure 1. Western ice sheet snowfall totals are 30%-70% of normal. Brown areas have less than ‘normal’ precipitation. Blue/purple areas are anomalously ‘wet’. The precipitation anomalies are calculated from ‘re-analyses’ data after Kalnay et al. (1996).

Multiple melt factors combine to increase the odds of more melt water runoff from the ice sheet during the 2013 melt season:

  1. less ‘cold content’ of snow to melt away (ablate) for a given energy input before bare ice is exposed;
  2. a longer period of exposed darker bare ice, in this case weeks earlier bare ice exposure is likely unless a big snow dump before or during the coming warm season;
  3. Less snow leads to a smaller refreezing capacity in the lower accumulation area. Thanks Robert Fausto of GEUS for reminding me of this one.
  4. a possible higher concentration of light absorbing impurities per unit volume of snow, assuming that the impurities are deposited whether or not it snows.

This pattern results from a persistent atmospheric anomaly, blocking cold air transport southward along west Greenland, producing relatively warm temperatures there while northwestern Europe has had a cold winter (Figure 2).

Figure 2. The data after Kalnay et al. (1996) indicate tendencies toward offshore flow over western Greenland, opposite for what is needed to produce normal snowfall. 

The precipitation anomaly is manifesting in abnormally low land and ice sheet reflectivity (albedo) (Figure 3).

Figure 3. April 2013 surface albedo (a.k.a. reflectivity) anomaly. Substantially lower albedo anomalies on land are due to the dearth of snow revealing a much darker underlying tundra. The red areas across the northern 1/3 of Greenland are uncertain due to low solar illumination angles.

Low snowfall anomalies precondition Greenland ice for enhanced melt (Mote, 2003; Box et al. 2005; 2012), especially for the western ice sheet where the snowfall amounts are less than over the east.

From 20 March – 20 April, the snow drought drove ice sheet reflectivity well below values in 13 years of (NASA MODIS sensor) satellite observations since 2000 (Figure 4). Negative North Atlantic Oscillation (NAO) has promoted Greenland heating, melting and snow drought for now 6 summers in a row (Tedesco et al. 2013; Fettweis et al. 2013). Negative late winter NAO packs a similar punch. Negative NAO has prevailed much of the past decade and is largely to blame for Greenland’s astonishing melt increase. Whether negative NAO is promoted by an earlier loss of snow on land and declining Arctic sea ice area is something I’ve been wondering about.

Figure 4. Greenland ice sheet (land excluded) reflectivity or albedo updated after Box et al. (2012).

Then the weather flipped and ice sheet reflectivity rebounded toward normal values in the latest 10 days (Figure 4). Ice sheet reflectivity and accumulated precipitation remains lower than average for the year to date through 1 May (not shown), it therefore remains more likely than not that we’ll see a big melt in 2013.


  • Box, J.E., L. Yang, J. Rogers, D. Bromwich, L.-S. Bai, K. Steffen, J.C. Stroeve, and S.-H. Wang, 2005: Extreme precipitation events over Greenland: Consequences to ice sheet mass balance. Preprints, Eighth Conf. on Polar Meteorology and Oceanography, San Diego, CA, Amer. Meteor. Soc., CD-ROM, 5.2. PDF
  • Box, J. E., X. Fettweis, J.C. Stroeve, M. Tedesco, D.K. Hall, and K. Steffen: Greenland ice sheet albedo feedback: thermodynamics and atmospheric drivers, The Cryosphere, 6, 821-839, doi:10.5194/tc-6-821-2012, 2012. open access
  • Fausto, Robert, provided point 3 above.
  • Fettweis, X., Hanna, E., Lang, C., Belleflamme, A., Erpicum, M., and Gallée, H.: Brief communication “Important role of the mid-tropospheric atmospheric circulation in the recent surface melt increase over the Greenland ice sheet”, The Cryosphere, 7, 241-248, doi:10.5194/tc-7-241-2013, 2013.
  • Mote, T., 2003: Estimation of runoff rates, mass balance and elevation changes on the Greenland ice sheet from passive microwave observations. Journal of Geophysical Research-Atmospheres, 108, 4056. DOI
  • Kalnay et al.,The NCEP/NCAR 40-year reanalysis project, Bull. Amer. Meteor. Soc., 77, 437-470, 1996.
    • The data are constrained by measurements from weather stations, weather balloons, ships, aircraft, and satellites. I total precipitation for 1 January – 30 April, 20130 and difference these with the average total for the same interval over each year in the most recent 30-year ‘climate normal’ spanning 1981-2010.
  • Tedesco, M., Fettweis, X., Mote, T., Wahr, J., Alexander, P., Box, J. E., and Wouters, B.: Evidence and analysis of 2012 Greenland records from spaceborne observations, a regional climate model and reanalysis data, The Cryosphere, 7, 615-630, doi:10.5194/tc-7-615-2013, 2013. open access