Estimates of future warming‐induced methane emissions from hydrate offshore west Svalbard for a range of climate models

Abstract

Methane hydrate close to the hydrate stability limit in seafloor sediment could represent an important source of methane to the oceans and atmosphere as the oceans warm. We investigate the extent to which patterns of past and future ocean-temperature fluctuations influence hydrate stability in a region offshore West Svalbard where active gas venting has been observed. We model the transient behavior of the gas hydrate stability zone at 400-500 m water depth (mwd) in response to past temperature changes inferred from historical measurements and proxy data and we model future changes predicted by seven climate models and two climate-forcing scenarios (Representative Concentration Pathways RCPs 2.6 and 8.5). We show that over the past 2000 year, a combination of annual and decadal temperature fluctuations could have triggered multiple hydrate-sourced methane emissions from seabed shallower than 400 mwd during episodes when the multidecadal average temperature was similar to that over the last century (;2.68C). These temperature fluctuations can explain current methane emissions at 400 mwd, but decades to centuries of ocean warming are required to generate emissions in water deeper than 420 m. In the venting area, future methane emissions are relatively insensitive to the choice of climate model and RCP scenario until 2050 year, but are more sensitive to the RCP scenario after 2050 year. By 2100 CE, we estimate an ocean uptake of 97-1050 TgC from marine Arctic hydrate-sourced methane emissions, which is 0.06-0.67% of the ocean uptake from anthropogenic CO 2 emissions for the period 1750-2011.

Resistivity image beneath an area of active methane seeps in the west Svalbard continental slope

“…These estimates are presented assuming that the seafloor profile for Line 1 and Line 2 are representative for the entire Area 3 between 400 and 430 m water depth. The amount of carbon available for dissociation from hydrates in our estimates is an order of magnitude higher than the amount Marín-Moreno et al (2015a) predicts may release in the next 100 yr. The model estimates of Marín-Moreno et al (2015a) suggest only a part of the hydrate in the vulnerable zone will dissociate and some of the methane from dissociation will stay in the seabed as free gas.…”

“…Modelling studies suggest that the GHSZ beneath 400-430 m water depth in the study area is vulnerable to increase in bottom water temperatures within the next century (Marín-Moreno et al 2013, 2015a. Extrapolating the inferred average hydrate saturation for the seafloor depths of 400-430 m (15-45 per cent between 9 Figure 14.…”

“…Since the presence of glacial sediments at the upper slope is likely to cause reduction in porosity Sarkar et al 2012), gas and hydrate saturations for this area were also estimated using an average porosity of 35 per cent, similar to that used by Marín-Moreno et al (2013) and Marín-Moreno et al (2015a). Bottom water temperature variations in the area are available from the DASI CTD for 50 m above the seafloor.…”

“…Since the methane seeps must be fed either by shallow dissociating hydrate or from a free gas reservoir beneath the seafloor, improved estimates of this methane inventory are crucial. Hydrate and gas saturation estimates also provide input to models predicting the future response of the subsurface methane to ocean temperature changes (Marín-Moreno et al 2013, 2015a.…”

Paleo‐methane emissions recorded in foraminifera near the landward limit of the gas hydrate stability zone offshore western Svalbard

“…However, dating of authigenic carbonates at sites of present‐day seepage indicates that the seeps have been active for at least 3000 years before present (BP) [ Berndt et al ., ]. Seasonal and decadal temperature variability, superimposed on longer term trends, are expected to have produced cycles of methane accumulation as gas hydrate in shallow sediments during colder conditions followed by gas release as temperature increases [ Berndt et al ., ; Marín‐Moreno et al ., ].…”