USGS visual identity and link to main Web site

Methane in Coastal Sea Water, Sea Ice, and Bottom Sediments, Beaufort Sea, Alaska

U.S. Geological Survey Open-File Report 95-70

by Thomas D. Lorenson and Keith A. Kvenvolden

This report is preliminary and has not been reviewed for conformity with U.S. Geological Survey editorial standards (or with the North American Stratigraphic Code). Any use of trade, product, or firm names is for descriptive purposes only and does not imply endorsement by the U.S Government.

Formal metadata describing these data are available.



This report summarizes data acquired from 1990 to 1994 for the gas-hydrate portion of the USGS project "Permafrost and gas hydrate as possible sources of methane" of the USGS Global Change and Climate History program. The objective of this project has been to test the hypothesis that gas hydrate deposits of the Beaufort Sea continental shelf are destabilized by the ~10C temperature increase that has resulted from the Holocene transgression of the Arctic Ocean. To test this idea we have selected an area off the north coast of Alaska centered on Harrison Bay. We have measured the concentration of methane in surficial sediments, in the water column when ice is present and absent, and in seasonal sea ice. Our results show that more methane is present in the water when ice is present than when ice is absent, and that methane is also present within the ice itself, often at higher concentrations than in the water. Thus the Beaufort Sea shelf of Alaska is a seasonal source of methane. The primary source of this methane has not yet been defined, but gas hydrate is a reasonable candidate.


Methane is an important trace component of the atmosphere. Its concentration is increasing at a rate of about 0.9% per year (Khalil and Rasmussen, 1983; Steele et al., 1987; Blake and Rowland, 1988), although this rate has decreased recently (Steele, et al., 1992). Because methane is radiatively active, it is a "greenhouse" gas that has a global warming potential 20 times larger than an equivalent weight of carbon dioxide when integrated over a 100-yr span of time (Shine et al., 1990). The earth's atmosphere has a wide variety of sources and sinks for methane (Cicerone and Oremland, 1988), including gas hydrate. Gas hydrate exists in metastable equilibrium with its environment and is affected by changes in pressure and temperature. The amount of methane that is trapped in gas hydrate is perhaps 3,000 times the amount in the atmosphere. A large release of methane from this source could have a significant impact on atmospheric composition and thus on the radiative properties of the atmosphere that affect global climate (MacDonald, 1990).

Gas hydrate deposits of polar continental shelves are currently believed to be the most vulnerable to climate change (Kvenvolden, 1988). These areally extensive shelves, formerly exposed to very cold surface temperatures (yearly average -10 to -20C) have been and are being transgressed by a much warmer polar ocean (~0C). As a result, the polar shelf surface in water depths greater than 5 m, has experienced an approximately 10C or more temperature increase over the past 10,000 years. Although pressure on shelfal gas hydrate deposits has increased owing to a rise in sea level of about 100 m, this pressure increase which would stabilize the gas hydrate, is more than offset by the large temperature increase that destabilizes the gas hydrate. The amount of methane released by this process has been estimated to be about 4 Tg (1012 grams) year-1 (Kvenvolden, 1991) or about 1% of all current sources of atmospheric methane. If this suggestion and estimate are correct, then escape of methane from gas hydrate deposits of polar continental shelves should be observable. To test this idea, beginning in 1990 we conducted surveys of methane in sediment, water, and ice in an area of the Beaufort Sea continental shelf offshore from the north coast of Alaska. Previously, measurements of atmospheric methane concentrations at Barrow, Alaska, about 200 km NE of the study area, showed a seasonal increase during August through October, perhaps due in part to methane venting from Arctic shelfal waters (figure 1). This report summarizes our data collected from 1990 through 1994.

Figure 1

Mean monthly atmospheric methane concentration (ppb-v, parts per billion) measured at Barrow, Alaska from 1983-1993. Measurments were made by the National Oceanic and Atmospheric Administration (NOAA), Climate Monitoring and Diagnostics Laboratory (CMDL). A PostScript version is also available.

These data are from Dlugokencky et al. (1994) and can be obtained through anonymous ftp from


In our survey, we measured methane concentration in sea water under the ice on six primary transects from nearshore to the shelf break at water depths ranging from 2.5 to 88m from Cape Halkett east to Mikkelsen Bay, Alaska (figure 2). Other surface water samples were taken at various locations where water depths ranged from 2 to 20m within the area between Camden Bay and the Colville River Delta (figure 2). One oceanic surface water sample was also collected in April 1993 at 72 38' N. latitude . The surveys were undertaken during the last week of April and first week of May 1992, the middle of April 1993, the latter portion of April 1994, the second week of September 1993, and the later portion of August 1994. At each station in April and May, a 23-cm diameter hole was drilled through the 2-m thick first-year ice. Niskin bottles (1.7 l) were attached at intervals to a line lowered to the seafloor and subsequently tripped to collect water samples for methane determinations. In September 1993 and August 1994, when no ice was present, a boat was used to transit to stations. As each Niskin bottle was recovered, 100 cm3 of water was transferred to a 140 cm3 syringe. A heat lamp to prevent the water from freezing during April and May surveys. In the laboratory at the end of each day's field work, 40 cm3 of ultra-pure nitrogen were added to each of the syringes, and methane was extracted from the water into the nitrogen at room temperature by shaking the syringe, following a method modified from McAullife (1971). A portion of the resulting gas mixture was measured for methane content by gas chromatography. Multiple extractions were used on several samples to determine an empirical extraction efficiency coefficient which was utilized with the remainder of the samples that were extracted only once. Our analyzed gas standard contained 10 ppm 2% methane in nitrogen. Duplicate samples were analyzed for each water depth with an average standard deviation of 0.6 nM.

Figure 2

The study area showing six north-south transects with the water and ice station locations. A PostScript version is also available.

CTD (conductivity, temperature, depth) casts were made at each station. CTD data was recorded internally by a Seacat 19 CTD probe made by Sea Bird of Seattle Washington. The data was downloaded onto a computer where it was processed using Seasoft Version 4.016. The sub-programs LOOPEDIT, BINAVE, then DERIVE were utilized to process the raw data. The data were averaged into 0.5m intervals. The upcast and downcast were evaluated, and only one cast is plotted. In general the upcast was used because the sensors showed less fluctuation. The results are plotted for each station and appended with water methane concentrations in Appendix 1.

Digital data from the CTD casts are available online by individual cast as tab-delimited ASCII (plain text) files or in their entirety as a UNIX-style archive (tar format).

The methods utilized to measure nutrients in this study were adapted from older colorimetric procedures and are described in detail by Whitledge and others, (1981). A basic description of each method, conducted on a Technicon Autoanalyzer System, follows.

Phosphate is determined as phosphomolybdic acid which in its reduced form in the presence of antimony has an absorption maximum at 880 nm. The method is basically an automated version of the procedure of Murphy and Riley (1962).

Orthosilicic acid is determined by its reaction with molybdate in aqueous acidic solution to form silicomolybdic acid. In this procedure, which is basically that of Armstrong and others, (1967), stannous chloride is used to reduce silicomolybdic acid to the heteropolic acid which has an absorption maximum at 820 nm.

Nitrite is determined by the Greiss reaction in which sulfanilamide and N-(1-Napthyl)ethylenediamine dihydrochloride react with nitrite in aqueous acidic solution to form an intensely pink diazo dye with an absorption maximum at 540 nm (Bendschneider and Robinson, 1952). Nitrate, after it is reduced to nitrite by passage through a column of copperized cadmium filings, is determined in an identical manner to nitrite (Wood and others 1967). This analysis gives the sum of nitrate + nitrite, and nitrate is determined by difference.

Ammonium is determined by the Berthelot reaction in which hypochlorous acid and phenol react with ammonium in aqueous alkaline solution to form indophenol blue, an intensely blue chromophore with an absorption maximum at 637 nm. The method utilized is a modification of the procedure reported by Slawyk and MacIsaac (1972).

The methane content of sea ice was measured at 51 stations in 1993 and 1994. A 1m long by 7.6 cm diameter ice corer with an extension was used to core through the ice to sea water. The ice core was laid out on the sea ice, measured and cut into 10 cm sections. Selected intervals were chosen, then placed in 1-liter friction sealed cans equipped with 2 septa ports, sealed, and kept frozen. At the field laboratory, each sample was weighed, then the headspace within the can was purged with ultra pure nitrogen for 5 minutes at flow rates exceeding 200 ml/min. Tests on the methane content of the purged headspace with the frozen ice core inside yielded concentrations less than 0.1 ppm. The cans containing ice samples were placed in hot water baths and stabilized at approximately 20 C. Each can was shaken by hand for about 30 seconds to partition the dissolved methane between the water and the nitrogen headspace. Next, a syringe containing 30 ml of ultra pure grade nitrogen was injected into the can, then 30 ml of the resulting mixture of headspace and gas was removed and analyzed for methane content in the same manner as described for sea water samples. The volume of headspace in the can was determined by the weight of the ice sample, assuming that once melted the water occupied a volume of 1 ml/gm.

Sediment samples were taken in May 1992 operating from the ice canopy. A hole was augured in the ice, through which a 45 x 7.5 cm core barrel with an auger cutting head was lowered. The core barrel was attached to 1.5-m long connecting stems in series up to a maximum length of 15.5 m. Once on the ocean bottom, the core barrel was rotated by hand with the aid of a T-bar until no further penetration was achieved. Upon retrieval, the liner containing sediment was extruded onto the ice. The least disturbed 10-cm long (a volume of about 450 ml wet sediment) section of the core was chosen and placed in a septa-equipped 1-liter sample can. Seawater was added to the brim of the can, then 200 ml of water was removed, creating a 200 ml headspace. About 2-3 grams of sodium azide was added as a biocide. The can was sealed, then purged with helium at a flow rate estimated to be greater then 300 ml/min. for 5 minutes. The samples were kept frozen until analysis at our Menlo Park laboratories.


Stations in water depths from 2.5 to 88m were established on six north- south transect lines as shown in figure 2. Three digit, abbreviated station numbers are shown on the transect lines. Many stations were reoccupied on subsequent surveys. Each station is given 3 additional numbers making a total of six digits per station. The first two refer to the year, the third to the month of sample collection, the fourth to the transect number, and the last two represent the approximate water depth at each station. For example, station 925150 can be read as 1992, May, Transect 1, at 50 meters water depth. When the same station was reoccupied on a subsequent survey only the first 3 digits change denoting a different sampling date; i.e., 934150 was completed in April, 1993. Stations off transect lines are given a 0 in the fourth digit place.

Station locations are available in a table in tab-delimited ASCII, with the station locations encoded as decimal degrees north latitude and west longitude.

Water Methane Concentrations

Water methane concentrations and methane oxidation rates are listed in Table 1, and are illustrated by graphs of methane vs. water depth at each transect in Appendix 2. Measurements of methane in ice-covered sea water were made in May 1992, April 1993, and April 1994. The results are compared with measurements taken in ice-free water during September 1993 and August 1994.

Analyses of 143 samples from ice-covered waters demonstrate that methane concentrations in the water column are supersaturated with respect to the equilibrium solubility of methane with the atmosphere (~4 nM). Maximum methane concentrations at a given station ranged from 12 to 275 nM and were typically about 15 to 25 nM throughout the water column. The highest methane concentrations were found near the bottom. Maximum methane concentrations are given in map view for each survey in Appendix 3.

Methane concentrations from ice-free water were determined in 103 samples. Maximum methane concentrations at a given station ranged from 5 to 44 nM with one anomalous measurement of 148.6 nM at station 948120, 18m water depth. Anomalous methane concentrations (33.7-94.4 nM) were also observed at this same station, 925120 throughout the water column in May, 1992. Typically, methane concentrations were at a minimum at the surface, remained nearly constant or increased with depth to concentrations ranging from 9 to 56.4 nM, excluding the 148.6 nM anomaly mentioned above. Methane concentrations in the upper 3m of ice-free water never exceeded 18 nM and were commonly between 6 and 14 nM.

A limited number of measurements of the carbon isotopic composition of methane were made at the University of Hawaii. The results shown in Table 1 reveal a wide range of values from -31.1 to -80.5 o/oo.

Methane Oxidation rates were determined in selected samples (Table 1) and are given graphically in Appendix 4. The rates during ice covered times were higher than most measurements made worldwide. In ice free periods measured to date (September, 1993), the oxidation rates were undetectable. Thus a seasonality exists regarding the oxidation of methane by bacteria living in the water column.

Table 1

Table 1 is provided in tab-delimited ASCII format. Tab characters are used to separate cells along each row. This includes the column headings, so when viewed in a WWW browser, the column headings won't line up with their data. But when imported into a spreadsheet or database program, the headings will line up correctly with their data.

Column headings:

Nutrients in water

The nutrients, phosphate (PO4), silica (Si), ammonium (NH4), nitrite (NO2), and nitrate (NO3) were measured at many stations throughout the study period. At this time only the data from the May 1992 survey is listed in Table 2.

Table 2

Table 2 is provided in tab-delimited ASCII format. Tab characters are used to separate cells along each row. This includes the column headings, so when viewed in a WWW browser, the column headings won't line up with their data. But when imported into a spreadsheet or database program, the headings will line up correctly with their data.

Column headings:

Methane in ice cores

We measured the methane content of samples from the ice canopy in April 1993 and 1994. Typically coastal sea ice inshore from 15m water depth initially forms as a mobile mass of frazil and brash ice a few decimeters thick. This mass gradually solidifies into granular "fast" or stationary ice by late fall. Subsequently the fast ice grows to a thickness of about 2m over a fixed point on the seafloor. Ice seaward of about 15m water depth (the seasonal ice pack) is formed in a similar manner but is subject to intense deformation and is mobile throughout the winter. Thus, fast ice is most likely to trap methane in transit from sea water to the atmosphere at a fixed location, whereas offshore ice most likely traps methane from non-fixed locations as it thickens.

We sampled fast ice and seasonal pack ice at 36 locations along the coast from Harrison Bay to Camden Bay in water depths ranging from 1.8 to 57 m. Maximum methane concentrations are given in Appendix 5 and the complete results are listed in Table 3. Methane concentration profiles along transects are given in Appendix 6.

Methane concentration profiles through the ice within the upper 50 cm interval and from water depths less than about 10m contained the highest concentrations of methane (7 to 1260 nM), with typical values of about 30 to 40 nM. In the ice interval from 50 to 200 cm, the methane concentration ranges from 5 to 286 nM, with typical values of about 15 nM.

Ice that formed in water depths greater than 15m shows the same trend of decreasing methane with increasing depth in the ice core; however, the methane concentrations were less varied and were always lower. Within the 0 to 50 cm interval methane concentrations ranged from 8 to 14 nM and from 5 to 14 nM within the 50 to 200 cm interval. Measurements of the carbon isotopic composition of methane in ice are shown in Table 3 and reveal a wide range of values from -52.1 to -83.4 o/oo.

Table 3

Table 3 is provided in tab-delimited ASCII format. Tab characters are used to separate cells along each row. This includes the column headings, so when viewed in a WWW browser, the column headings won't line up with their data. But when imported into a spreadsheet or database program, the headings will line up correctly with their data.

Column headings:

Methane in sediments

In May 1990, our initial study of gas sources in Harrison Bay concentrated on surficial sediments. It was thought that if methane from destabilized gas hydrates were accumulating and migrating, the gas would transit through the sea floor. As a guide to sampling locations we reviewed the work of Neave and Sellman (1982), who seismically mapped subsea (drowned) permafrost and shallow gas anomalies. These mapped areas and our measured methane concentrations in sediment are shown in figure 3, and the data are available in digital form in table 4. Our working model of gas hydrate occurrence in permafrost (Kvenvolden, 1988) requires that gas hydrate and permafrost coexist. Holocene sea level rise has, within the last 10,000 years, flooded the area now know as Harrison Bay. About a 10C temperature rise over this area has initiated a downward migrating thermal pulse capable of melting permafrost and destabilizing gas hydrate. With the additional thermal insulation of the ocean, the geothermal gradient creeps upward and causes an upward migration of warmer temperatures (Lachenbruch and others, 1988). Ostercamp and Fei (1994) have shown that the combined effect of the temperature increases above and below the permafrost should have begun to melt the outer portions of the subsea permafrost, corresponding to approximately the 20m isobath.

Table 4

Table 4 is provided in tab-delimited ASCII format. Tab characters are used to separate cells along each row. This includes the column headings, so when viewed in a WWW browser, the column headings won't line up with their data. But when imported into a spreadsheet or database program, the headings will line up correctly with their data.

Column headings:

Figure 3

Methane concentrations in sediment (l/l, microliters methane per liter wet sediment) measured in May, 1992. Shallow gas anomalies taken from Neave and Sellman, (1982). The contour interval is 2 meters. A PostScript version is also available.

Methane concentrations range from 2.2 to 10.1 ml/l wet sediment. There is no distinctive pattern of methane concentration. However the highest concentrations were recorded within the confines of one mapped shallow gas anomaly. The highest concentrations are found near the Colville River Delta and range from 4.6 to 10.1 ml/l. The lowest concentrations of methane (3.8 and 2.2 ml/l) were found offshore at Weller Bank in clean sand. This low concentration is perhaps the result of methane escaping more readily from the clean sand rather than silt or clay which comprise the bulk content of the rest of the samples.


This report documents the concentrations of methane in sediment, water, and sea ice in an area of the Beaufort Sea continental shelf offshore northern Alaska. Data were collected during field surveys conducted from 1990 to 1994. In addition to methane concentrations, information regarding methane oxidation rates, nutrients, salinity, and temperature of the water column is also tabulated. In general, more methane is present in the water when ice forms a seal over the water column than when ice is absent. Methane is also present in ice often in concentrations exceeding those in water. Fast ice forming in place over water depths less than about 10m can contain methane concentrations an order of magnitude higher than that of the underlying water. Measurements of the carbon isotopic composition of methane and ice were initiated in 1994 as an attempt to define the source or sources of methane. These data represent the first ever collected is such a setting. The data thus far demonstrate a wide range of values from -31.6 to -83.4 o/oo indicating that there a may be complex variety of sources.


The project has been conducted in collaboration with the University of Washington and the University of Hawaii. We thank (1) M.D. Lilley, E.J. Olson, and E. McLaughlin, School of Oceanography, University of Washington, for the information on methane concentrations, methane oxidation rates, and nutrient concentrations in seawater; (2) B.N. Popp, F. J. Sansone, and T. Rust, SOEST, University of Hawaii, for the isotopic analyses of methane. We are grateful to (1) P.W. Barnes, U.S. Geological Survey, for guidence in Arctic operations, and (2) E. Reimnitz for help in collecting sediment samples for gas analyses. This work is supported by the USGS Global Change and Climate History Program.


Technical contacts

Thomas D. Lorenson
Mail Stop 999
U.S. Geological Survey
345 Middlefield Road
Menlo Park, CA 94025
Tel: (415) 354-3094
FAX: (415) 354-3191

Keith A. Kvenvolden
Mail Stop 999
U.S. Geological Survey
345 Middlefield Road
Menlo Park, CA 94025
Tel: (415) 354-3213
FAX: (415) 354-3191


Information here is presented as PostScript figures only. The files include sufficient information to permit printing on any standard PostScript-compatible device; the maps are also compatible with Adobe Illustrator version 5.0 for the Macintosh.

  1. Water Temperature - Salinity - Methane concentrations plotted with water depth for each station
  2. Methane concentrations in water plotted with water depths grouped by transects from each survey
  3. Map view of maximum methane concentrations (nM) in water for each survey
  4. Methane oxidation rates of water plotted with water depth grouped in transects conducted in each survey
  5. Map view of maximum methane concentrations in ice for each survey
  6. Methane concentrations of ice plotted with ice depth, grouped in transects

This page is <URL:>
Maintained by Eastern Publications Group Web Team
Last updated 27-Feb-2001 (krw)