Ice core drilling and subglacial lake studies on the temperate ice caps in Iceland

Iceland astride the Mid-Atlantic ridge

Lecture # 36 Astrobiology Winter School Ice drilling projects in Iceland University of Hawaii, Jan. 2005 Thorsteinn Thorsteinsson ([email protected]) National Energy Authority Ice cover during the late glacial period. Combination of hot spot volcanism, spreading on the Mid-Atlantic ridge and glaciation leads to unusual geology. From Thordarson and Höskuldsson (2003).

Subglacial volcanism beneath an ice sheet creates unusual landforms:

Tuyas (tablemountains) and ridges.

Herðubreið, N-Iceland Submarine eruptions create similar formations.

Biologicial colonization monitored from the beginning Surtsey eruption, S. of Iceland, 1963-1967. Recent subglacial eruption in Grímsvötn, Vatnajökull ice cap. Jökulsárgljúfur canyons, N. Iceland: Formed in a catastrophic flood from Vatnajökull 2500 years ago. Icelandic analogs to hillside gullies on Mars

Mars Mars Iceland

Iceland

Iceland Temperate ice caps Hofsjökull, 890 km2 cover >10% of Iceland

Vatnajökull 2 Langjökull 8000 km 2 920 km Max. thickness: 950 m

Mýrdalsjökull 550 km2

- Dynamic ice caps in a maritime climate. - High accumulation rates (2-4 m w.eq. yr-1) - Extensive melting during summer, except at highest elevations (> 1800 m) - Mass balance of ice caps negative since 1995

1972: A 415 m ice core was drilled on the NW-part of Vatnajökull Activities not continued at that time, drill was discarded. Mass balance of Hofsjökull Ice Cap

4

2

0

Winter balance -2 Summer balance H2O Net balance m -4 Cumulative mass balance -6

-8

-10 1987 1992 1997 2002 Climate, Water, Energy: A research project investigating the effect of climate change on energy production in the Nordic countries

Temperature increase 1990-2050 °C

CWE uses IPCC data to create scenarios for temperature and precipitation change in the Nordic region in the future Modelled change in volume of Hofsjökull ice cap 2000-2002, Modelled glacial meltwater using four different dT and dP discharge (run-off) 2000-2200. scenarios. SINCE 2001: NEW EFFORTS IN ICE CORING AND DRILL DESIGN

100 m ice core drilled at Hofsjökull summit,1790 m above sea level, using a German-built shallow drill. - What kind of drill design is suitable for coring in temperate ice? - Can annual layers be identified in the cores? - Can climate records be retrieved from the temperate ice caps?

16-19 m 40- 43 m 73-76 m Polar ice:

Temperature below melting point, borehole dry (Deep coring: Liquid dumped in hole, to prevent closure)

Temperate ice:

At melting point throughout! Meltwater (and rain) seeps down into snow and firn, and is present in small veins on crystal boundaries below the firn-ice boundary! Water table in ice cap! 1.0

0.9

] 3

3 Firn-ice boundary ( ! = 0.83 g/cm ) 0.8 Ice becomes impermeable [g/cm 0.7

Density 0.6 Drill hole water-filled 0.5 below firn-ice boundary!

0.4 0 20 40 60 80 100

Depth [m]

Hofsjökull – density profile Shallow ice core drill built at the Alfred Wegener Institute in Germany 1 m

Antitorque Motor Hatches Core barrel

Dry hole: Chips fall through hatches and are collected in upper part of barrel. Core length ~ 1 m.

Drill head This system does not work well in a water-filled hole, where:

- Motor must be water-protected - Travel time up & down in water column must be > 0.5 m/s - Chips chamber must be designed in a different way A different design was tested with success on the ice shelf covering the Grímsvötn subglacial lake in 2002

1.6 Core length drilled in each run Grímsvötn water table 1.4 Hofsjökull 2001

Grímsvötn 2002 1.2 Hofsjökull water table

(m) 1.0

length 0.8

Core 0.6

0.4

0.2

0.0 0 20 40 60 80 100 120

Depth (m) Core length in each drilling run increases with new system! 18O/16O - 18O/16O 18 sample SMOW δ O = x 1000 ‰ 18 16 O/ OSMOW

18 LOCATION ALT Tm δ O

SMOW 0.0 ‰ Reykjavík 14 m 4.4 °C - 8.0 ‰ Hofsjökull, Iceland 1800 m ~ - 6 °C - 12.9 ‰ Greenland summit 3200 m - 32 °C - 35.0 ‰ Hofsjökull δ18O profile (0-20 m)

0 2001

5 2000

Summer [m] 1999 surface 10 Depth

1998

15

1997 Summer surfaces determined from:

20 1996 - annual mass balance measurements -16 -14 -12 -10 18 - dust measurements on core ! O [‰]

Identification of annual layers from δ18O data not straightforward! 0 5 10 15 20 0 0 0 0 0 0 0

5 5 5 5 5

10 10 10 10 10

15 15 15 15 15

20 20 20 20 20 20 20

0 200 400 600 800 0 400 800 1200 0 100 200 300 0 50 100 150 200 0 50 100 150 200 + - 2- ++ - Na Cl SO4 Ca NO3

Marine aerosol dominates. 40 40 Washout occurs during first melting season!

60 60

80 80

100 100 0 20 40 60 80 100 0 5 10 15 20 Meltlayer % DEP conductivityx10 -6 (µS/m) Lava fields and sandy deserts in the interior are snow covered in winter, but snowfree 2-3 months during summer:

Windblown dust covers ice caps in late summer!

(Ó. Arnalds et al. 1997) Temperature, stratigraphy and dust concentration in a 15 m snow core at drill site May 2003

0 0 0

2 2 F 2 1

2 4 4 4

3 F 6 4 6 6 [m] 5 8 8 8 D_pi 6

F 10 710 10

8 2002 summer

12 12 12 G surface 9 F

G 14 1014 14

11 G -6 -4 -2 0 0 5 10 15 20 0 5 10 15 20 H18 F 12 F 13 G 2001 summer 14 G surface 15

T [°C] Ryk 10 cm [ppmv] Ryk 20 cm [ppmv]

Dust [ppmv] 40 Hekla 17. janúar 1991

1990 Dustpeaks allow identification 45 1989 of annual layers.

Volcanic ash layers provide 50 1988 reference horizons.

1987

55 1986

1985 D_pi (m)

60 1984

1983 ?

65 1982

1981

70 Hekla 17. ágúst 1980 0 2 4 6 8 10

Grímsvötn eruption November 2 2004

Dust concentration (ppmv) Hofsjökull 2001: Dust concentration and oxygen isotopes 0

20

40 D?pi (m)

60 Net accumulation from ice core Accumulation - yearly mass balance measurements Hveravellir precipitation 80

Comparison with the precipitation record from 100 0 1 2 3 4 5 Hveravellir weather station (650 m a.s.l.), 35 km

?ykkt árlags (m) west of drilling site.

Measured annual layer thicknesses.

Correction for density differences and plastic thinning of layers with depth yields net mass balance at drilling site 1970-2000. Hveravellir

-11.5 Mean annual temperature (Sept-Aug) - Hveravellir 0.0 core -12.0 mean ice -0.5

-12.5 temperature -1.0 Hofsjökull

- -13.0 -1.5 [‰] [ O -13.5 °C

18 -2.0 !

18 ] ! O values - Hofsjökull ice core -14.0 1975 1980 1985 1990 1995 2000

Year (September-August)

-12.0 18 ! O = 0.37 * T - 12.62 [‰] 18 -12.5 Mean annual values of δ O in ice core correlate with mean annual temperature -13.0 at Hveravellir.

2 O [‰] - Hofsjökull ice core -13.5 r = 0.29 18 !

-2.0 -1.5 -1.0 -0.5 0.0

Mean annual T (Sept-Aug), Hveravellir [ °C ] How many annual layers are preserved in the Icelandic ice caps?

Sandwich model – simple model that describes the thinning of annual layers with depth. Assumes vertical strain rate constant throughout ice sheet.

H H Change in t(y)= ln ë y annual layer H thickness (λ) with depth ! y y H H ë H

ë ë = y H y H

0 x λ λΗ Dansgaard-Johnsen model: More sophisticated model that does not assume constant vertical strain rate. Melting at the bed can be assumed in model y y

h

!v !v " y " y !y !y

Sandwich model: Dansgaard-Johnsen model: Vertical strain rate constant Vertical strain rate constant from surface to bedrock down to distance h above bed, (unrealistic in lower part) but decreases linearly with depth to 0 below h. Age vs depth relationship according to the Dansgaard-Johnsen model

' ' 2a $ 1 $ % %kh + k 0 "H ( kh 2 + a " % " 0 & & 2a # 1 # 1 % & k # 2 " $ 0 2 ! t ln h y H $kh + k !H ' kh + a0 = % " ! ! $ $ k ! 2 ! 2a0 % ' 2a $ 1 " 1 % " kh + k %kh + k 0 "y'( kh 2 + a t' = ln$ ! k % % " 0 " k 2 2a0 $ & 2a # 1 ! & & # # kh + k $kh + k 0 !h ' kh 2 + a k $ $ ! 0 ! ' $ % k " 2 % " % " 2 1 1 t ' $% " t' 0 y h = % "% ( " + ! ! & k # 2a0 2a0 % y'+ h + " & k k #

Results for Hofsjökull summit (H = 300 m, λH = 3.6 m ice)

x

500 year record could be retrieved, and annual layers still detectable at 270 m! Bárðarbunga, NW Vatnajökull (2000 m a.s.l.)

Model results at 700 m depth: 1100 years old ice Annual layers 17 cm thick

0 0

Bár!arbunga 200 200

400 400 Depth (m) Depth (m)

600 600 Bár!arbunga

800 800 0 400 800 1200 0.0 0.5 1.0 1.5 2.0 2.5 Age (yr) Annual layer thickness (m) - New drill being built in Iceland, operational in 2005. The drill could be made available for coring efforts on temperate or polythermal ice caps elsewhere (Svalbard, Scandinavia, Patagonia)

- Ice cores from temperate ice caps in Iceland can be accurately dated

- Proxy records for precipitation and temperature can be obtained

- A 500 year record could be retrieved with a 300 m drilling at the summit of Hofsjökull

- A record covering historical time in Iceland (870 AD – present) could be retrieved with a deep drilling (>700 m) in the ice-filled caldera at Bárðarbunga, Vatnajökull ice cap *

Vatnajökull (8000 km2) The Grímsvötn volcanic caldera beneath the Vatnajökull ice cap:

Powerful geothermal system, frequent eruptions. Steady 115 m kjarni melting of ice from below, up * to 300 m thick ice shelf covers a subglacial lake that empties out in jökulhlaups every few years. Drilling site in Grímsvötn 2002 Lake Vostok: The largest of ~70 subglacial lakes beneath the Antarctic ice cap

Flat ice sheet surface above lake

Plans to sample the lake under development.

Lake Vostok possibly hosts microbial life, isolated from other life on Earth over millions of years.

Webpage: “Subglacial Antarctic Lake Exploration: http://salegos-scar.montana.edu/ Vostok core reaches 420,000 years back in time.

Drilling was stopped 150 m above the subglacial lake, after drilling 70 m into ice accreted to the base of the ice sheet. Microbes were found in the accreted ice.

Kerosene was used as a drilling fluid, and hence biological contamination cannot be ruled out. Recent hypothesis (P.B. Price):

Microscopic liquid veins on the boundary between three crystals in deep Antarctic ice are microbial habitats!

Psychrophilic bacteria can move in the veins and obtain energy and carbon from ions in solution. *

Vatnajökull (8000 km2) Grímsvötn 2002:

Core drilling to 115 m and hot water drilling through 280 m thick ice shelf to 115 m kjarni sample the subglacial lake * for a biological study.

Drilling site in Grímsvötn 2002 0 Dýpi 2000 Hofsjökull ice core Hekla 2000 1 cm (m) 1999 1998 33 annual layers identified 1997 with dust measurements 2020 1996

No dustpeak 1995 1994? 43 m 1993 1992 4040 1991 Hekla 1991 1990 65 m 1989 1988 1987 Rapid crystal growth! 1986 1985 30 year old crystals at 100 m 6060 1984 No dustpeak 1983? equal in diameter to 100,000 1982 year old crystals at 2850 m 1981 depth in the Greenland ice 1979 Hekla 1980 sheet. 1978 1977 81 m 80 1976 80 1975 1974 1973 1972 1971 1970 1969 100100 100 m 0 5 10 15 Dust concentration (ppmv) Growth rates in top 100 m: Comparison of Hofsjökull / Vatnajökull data with polar locations.

East Antarctica (T = - 55 °C): ~ 10-5 mm2/yr

Greenland summit (T = -32 °C): ~ 10-3 mm2/yr Grímsvötn:

Hofsjökull/Vatnajökull (T = 0 °C): ~ 102 mm2/yr

The effect of high concentrations of dust on crystal growth appears similar in both polar and temperate 67 m ice!

Within dust layer

95 m

Dust: 1 – 50 µm 98 m 2002: Hot water drilling into the Grímsvötn subglacial lake

Investigators: Eric Gaidos (U. of Hawaii) Brian Lanoil (UC Riverside) Th. Thorsteinsson (NEA / U. of Iceland) Grímsvötn Caldera Lake

• Altitude = 1400 m • Volume of ~1-5 km3 • Depth of 150 m • Ice cover of 250 m • Glacial drainage basin of 300 km2 • 4250 MW over 100 km2 • Episodic drainage (1 km3 @ 104 m3 sec-1)

Project Goals

1) Determine if there is an active microbial community in the lake 2) Identify dominant lineages/species using molecular techniques 3) Assay community potential to fix carbon and nitrogen

DAPI CELL COUNTS

M a i n T i t l e

1 . 0 E + 0 8

I c e C o r e L a k e s e d i m e n t A 1 . 0 E + 0 7 S n o w L a k e s e d i m e n t B

B o r e h o l e A L a k e s e d i m e n t C

B o r e h o l e B

B o r e h o l e B 2 1 . 0 E + 0 6

B o r e h o l e C a f t e r

p e n e t r a t i o n

L a k e W a t e r A c e l l s / m l 1 . 0 E + 0 5 L a k e W a t e r B

L a k e W a t e r C

J a c u z z i

J a c u z z i V e n t 1 . 0 E + 0 4

1 . 0 E + 0 3

R o w 6 1 . 0 E + 1 0

1 . 0 E + 0 9

1 . 0 E + 0 8

1 . 0 E + 0 7

c e l l s / g r a m

1 . 0 E + 0 6

1 . 0 E + 0 5

1 . 0 E + 0 4

G r i s m - O c e a n O c e a n L a c u s - F o r e s t F o r e s t M a r i n e A g s o i l s

v o t n > 2 0 0 < 2 0 0 t r i n e s o i l > 1 s o i l < 1 s e d i m e

m m m m n t s source: Whitman et al. (1998) Main results of 2002 pilot study on Grímsvötn subglacial lake:

- Microorganisms in lake water and bottom sediment - DNA analysis indicates that the lake community is distinct from communities in the snow and ice - Gene sequences highly similar to known psychrophilic organisms

Paper on results can be downloaded at: www.liebertpub.com/ast Future work:

We hope to carry out a 4 year program sampling the Grímsvötn lake and the nearby Skaftá cauldrons (beneath 500 m thick ice cover).

A new hot water drill has been built in Iceland for this purpose.