Introduction

Methane in marine sediments forms in the final stage of microbial degradation of organic matter when all sulphate in sediment porewater has been exhausted. Continued formation of methane below the sulphate zone may lead to the formation of free gas if methane solubility is exceeded. This situation is fairly common for organic-rich muddy sediments on the continental shelf and in estuaries in particular (Whiticar 2002). Once formed either dissolved in porewater or as free gas, methane will start to move upwards towards the seafloor by diffusion or by advection. Most methane will, however, be oxidised at the sulphate-methane transition zone concurrent with sulphate reduction (Iversen and Jørgensen 1985) before it reaches the seafloor. Yet, methane escapes from the seafloor in some cases, e.g. areas of coarse-grained sediments (Hovland and Judd 1988; Laier et al. 1992) where buoyancy of methane gas may overcome the capillary forces of sediment pore throats. The buoyancy of gas is, however, insufficient to permit movement of gas bubbles towards the seafloor in fine-grained sediments (Sills and Wheeler 1992), thereby preventing fast passage of massive methane through the sulphate zone in muddy sediments. The theoretical consideration on buoyancy of gas and capillary forces of sediment pore throats does not totally exclude gas escape from fine sediments (e.g. García-Gil et al. 2002), possibly due to sediment failure (Horseman et al. 1999; Boudreau et al. 2005).

Anaerobic methane oxidation appears to operate in reverse to methanogenesis, i.e. methane is converted to carbon dioxide and hydrogen by archaea in close concert with sulphate-reducing bacteria which keep hydrogen at a very low level (Boetius et al. 2000). The free energy of the process is very low, hardly able to support any biomass growth (Schink 1997). Therefore, in spite of its effectiveness, anaerobic methane oxidation is very sluggish compared to other microbiologically catalysed sediment processes, and may allow the broad coexistence of sulphate and methane at turnover times of months to years (Jørgensen et al. 2001). So, diffuse methane escape from the seabed may be possible in some areas, although this remains to be proved.

Methane escape may have serious implications for the local marine environment (e.g. Weeks et al. 2004) as well as for future climate, and therefore risk assessments and global budget estimates have been made (e.g. Fleischer et al. 2001; Judd 2004). In order to obtain reliable estimates, it is important to get as precise an overview of methane gas occurrences and their proximity to seabed as possible. Information on gas occurrences can be obtained by shallow seismic methods (Judd and Hovland 1992) but is not often reported in detail, as shallow gas mapping is not the primary goal of many shallow seismic surveys. Therefore, shallow seismic data from numerous surveys carried out for different purposes in the Skagerrak–western Baltic Sea region were re-examined in order to map shallow gas occurrences as accurately as possible. The data have been collected in a GIS contour map of depths to shallow gas including gas escape features, which will be presented in this paper.

Estimates of methane escape

Iversen (1995) estimates that some 87–99% of all methane produced in marine sediments is re-oxidised at the subsurface before reaching the seabed. Nonetheless, a vast literature exists on methane escape from the seabed because of the often interesting and important features associated with it, such as methane-derived authigenic carbonates (N.O. Jørgensen 1989b; Jensen et al. 1992) as well as rich autotrophic fauna (Dando et al. 1994a). Since methane is an important greenhouse gas (Lashof and Ahuja 1990), attempts have also been made to provide estimates of the relative contribution to the atmospheric methane budget from the marine environment. Estimates have been made regionally in (1) the Kattegat (Dando et al. 1994b), (2) the North Sea (Hovland and Judd 1988, 1992) and (3) the entire UK seas (Judd et al. 1997) as well as globally (Fleischer et al. 2001; Judd et al. 2002; Judd 2004). Estimates of the regional methane emission into the water column from the seafloor of shelf areas, based on seep occurrences, range over four orders of magnitude, from 2 × 103 g km−2 year−1 (Dando et al. 1994b) to 107 g km−2 year−1 (Hovland and Judd 1988, 1992). A recent estimate constrains this value for the North Sea to between 5 × 105 and 1.5 × 107 g km−2 year−1, and an emission from the continental shelf of the UK of 0.12–3.5 Tg C year−1 (Judd et al. 1997). The large discrepancy between the estimates emphasises the need for a better understanding of methane seeps and the regulation of methane emission.

GIS database of shallow gas and gas seeps

Previous reports on gas distribution in the Skagerrak–northern Kattegat region Fält (1982; van Weering 1982) emerged merely as by-products of Quaternary sediment mapping using shallow seismic methods. Therefore, the maps of Fält 1982 and van Weering (1982) are insufficient to locate areas of potential gas escape from the seabed. To obtain a map for this purpose, we constructed as a first step a GIS database on shallow gas occurrences in the Skagerrak–Baltic Sea region based on data from existing shallow seismic surveys. In this paper, we present the amount of data available for this purpose and the methods we used to construct depth contours of shallow gas in the region. The database also includes the location of known seeps or features known to be associated with seeps both from the literature and internal reports. Using this database, shallow gas depths may easily be compared with other relevant data to understand the geological controls of gas escapes within this region. Data on bathymetry, sediment type (Hermansen and Jensen 2000), depth of pre-Quaternary rock, organic matter (Jansen et al. 2003) and sediment accumulation rate already exist as GIS databases. The shallow gas GIS database may also serve as a tool for selecting target areas for studies of potential diffuse gas escape from the seabed. The paper will also briefly mention the work which has been done to allow even the inexperienced user to benefit from the data stored in the shallow gas database.

Shallow gas as a monitoring tool

Although the shallow gas GIS database was constructed to provide a better understanding of the geological controls of methane escape from the seabed, it may serve other purposes as well. Shallow gas depths may, if properly calibrated with porewater chemistry data, provide information on the state of condition of ecologically vulnerable areas. Such areas could be estuaries and bays subjected to long-term eutrophication due to increases in the discharge of nutrients from high-productivity farming. Occasionally, catastrophic events occur when shallow methane is escaping form the seabed, carrying with it toxic hydrogen sulphide (e.g. Weeks et al. 2004). To prevent this from happening in future, the Danish government decided that nutrient discharge should be reduced significantly and that the state of condition of the marine environment should be monitored carefully (Kronvang et al. 1993). The paper will present the example on how the shallow gas GIS database may be used to evaluate the state of condition of the Aarhus Bay area, which has been monitored for ecologically relevant parameters for many years.

Setting

The western Baltic Sea and Kattegat area form part of the North German and Danish sub-basins, which contain sedimentary deposits over 8,000 m thick in the central parts. Deposits from most geological periods since Cambrian times are present in the sediment record revealed by drilling, although uplift and erosion have occurred in certain areas, particularly during Cainozoic time. Deeper hydrocarbon accumulations have not been found except for some very minor ones off the north German coast in the western Baltic Sea. Shallow gas accumulations in the area are therefore presumed to be almost entirely of Holocene age, although some accumulations of Pleistocene age exist in the northern Kattegat region (N.O. Jørgensen 1990b; Laier et al. 1992). Shallow gas accumulations are often found in relatively thick fine-grained Holocene deposits in basins/depressions formed by glaciers, ice-streams or melt water rivers during the last ice age (Fält 1982; van Weering 1982). However, in order to be able to understand shallow gas distribution in greater detail, we must know the depositional history during the Holocene of the different areas in the region. Sedimentation during the Holocene was controlled by the interplay between changes in eustatic sea level and isostatic rebound, which resulted in changes in local sea level by as much as 50 m over the first 2,500 years after deglaciation in the Skagerrak-northern Kattegat region (Fält 1982). More information on the depositional history and present-day deposition/erosion may be found in the following reviews and studies for (1) the Skagerrak (Kuijpers et al. 1993; van Weering et al. 1993), (2) the northern Kattegat (Kuijpers et al. 1993; Lykke-Andersen et al. 1993) and the southern Kattegat (Lykke Andersen et al. 1993; Jensen et al. 2002), (3) Aarhus Bay (Jensen and Laier 2003 and references therein), (4) the Baltic Sea-Kattegat transition (Bennike et al. 2004) and (5) the western Baltic Sea (Björck 1995; Bennike and Jensen 1998; Andrén et al. 2000).

Present-day sedimentation rates based on the 210Pb method (van Weering et al. 1987; Lundqvist et al. 2003) differ strongly between the various areas, ranging from over 6 mm year−1 to essentially zero. Even higher sedimentation rates were indicated for a few stations in the Skagerrak but irregular variation of 210Pb with depth indicated that this may be due to sediment slumping. Areas of erosion have also been identified (Kuijpers et al. 1993), and erosion was invoked as an explanation for methane-derived authigenic carbonate structures rising up to 4 m above the seafloor (N.O. Jørgensen 1989b). The distribution of accumulation and non-accumulation areas relevant to this study may be inferred from the sediment type map in Fig. 1. Areas with mud and sandy mud can be classified as accumulation areas, glacial till and lag sediments are found in areas of erosion, and sand may represent areas of either deposition or erosion. A close-up of the sediment type map for the Aarhus Bay area is shown in Fig. 2. The thickness, up to 200 m, of the Holocene deposits across the northern Kattegat can be seen on the geological profile in Fig. 3, which was based on seismic profiles and the hydrocarbon exploration well Frederikshavn-1 (Lykke-Andersen 1992). The thickness of Holocene sediments is much less in the Aarhus Bay area, as can be seen on the geological profile of Fig. 4. This profile was constructed from shallow seismic profiles (boomer) together with results of vibro-coring (Jensen and Laier 2003). The high-resolution seismic survey permitted identification of four different depositional units for the Holocene (Fig. 4), which document changes in the depositional environment resulting from changes in the local seawater level. The thick Holocene deposits of the northern Kattegat may also consist of different units but these cannot be recognised using seismics which have a lower resolution compared to shallow seismic surveys. The depositional environment in the Skagerrak-northern Kattegat has changed with time during the Holocene, as has been documented by the radiocarbon dating of a large number of marine shells from a 200-m core of the Skagen-3 well (Conradsen and Heier-Nielsen 1995), drilled at the northernmost tip of Denmark. Results of an 8-m-long piston core in the southern Skagerrak also indicate fairly large changes in depositional environment over a relatively short time period (Hass 1993).

Fig. 1
figure 1

Simplified marine sediment map of the Skagerrak-western Baltic Sea (after Hermansen and Jensen 2000)

Full size image
Fig. 2
figure 2

Sediment map of the Aarhus Bay area (after Hermansen and Jensen 2000)

Full size image
Fig. 3
figure 3

Geological profile across the northern Kattegat (after Lykke-Andersen 1992; for location, see line FH-S in Fig. 10). The stratigraphy was constructed from the Frederikshavn-1 exploration well indicated on the left (Sorgenfrei and Buch 1964)

Full size image
Fig. 4
figure 4

Geological profile across Aarhus Bay (for location, see Fig. 6 line A-A′)

Full size image

Data sources and methods

A large number of shallow seismic surveys carried out for different purposes, applied as well as scientific, were available for the construction of the shallow gas depth-contour map of the Skagerrak-Baltic Sea region.

Surveys and instrumentation

Existing data

Tracks of the available surveys are indicated for two areas, the Skagerrak-northern Kattegat and Aarhus Bay in Figs. 5 and 6. The density of shallow seismic lines is approximately the same for the rest of the region. The shallow seismic equipment used for applied purposes, site surveys and the search for raw materials (sand and gravel) commonly consisted of a combination of three tools: a 3.5 kHz pinger, 0.6–2 kHz boomer and 100 kHz side scan sonar. Scientific surveys carried out in order to unravel the depositional history of a particular area or to map habitat areas used the same combination of tools or at least a combination of two different sub-bottom profilers, a high-resolution instrument and a lower-resolution instrument affording better penetration. The high-resolution instrument was one of the following: an 18 kHz profiler, 3.5 kHz pinger, 0.4–10 kHz X-star full spectrum sonar or 1–10 kHz chirp; the lower-resolution instrument was either a 0.6–2 kHz boomer or 800–1,200 Hz single channel sparker.

Fig. 5
figure 5

Tracks of shallow seismic lines used for the construction of the shallow gas map of the Skagerrak-northern Kattegat area. Thick grey lines are from Fält (1982), grey lines from GEOMAR Poseidon surveys, black lines from Danish surveys for raw materials (Larsen et al. 1986) and for habitat mapping (Larsen 1996)

Full size image
Fig. 6
figure 6

Tracks of shallow seismic lines used for construction of the shallow gas map of the Aarhus Bay area. Grey lines are from a Danish survey for raw material, black lines from a METROL survey (Jensen and Laier 2003). Line A-A′ indicates the location of the geological profile in Fig. 4. Circles Sampling stations for porewater analysis

Full size image

Surveys for applied purposes were carried out by GEUS and by commercial companies. Scientific surveys were carried out either by GEUS, often in collaboration with other research institutes, i.e. IOW in Warnemünde, GEOMAR in Kiel, and the University of Gothenburg, or by one of the latter three institutes alone. GEUS’s database holds digital tracks of most shallow seismic surveys of the region, except for the Poseidon surveys in the Skagerrak (Fig. 5), for which we obtained the tracks in digital form from GEOMAR to complete this work. Tracks in the Swedish part of the northern Kattegat were obtained from the thesis of Fält (1982). No acoustic data were obtained from the Swedish sector of the southern Kattegat, which covers the area with muddy sediments along the Swedish west coast (Fig. 1).

New data

In order to improve the quality of the gas depth-contour map, new shallow seismic data were acquired during recent cruises carried out within the framework of the EC METROL project. These cruises covered the gassy areas of the northern Kattegat (Fig. 5; METROL 2003), Aarhus Bay (Fig. 6; Jensen and Laier 2003) and the Baltic Sea: Bornholm and Arkona basins plus Mecklenburger Bay (METROL 2004). The acoustic instruments used during these surveys comprised a 0.4–10 kHz X-star full spectrum sonar, 1–10 kHz chirp, 0.6–2 kHz boomer and 800–1,200 Hz single channel sparker.

Navigation

Newer acoustic surveys used DGPS navigation, which in most cases was accurate to within 5–10 m. Older surveys, typically carried out before 1990, used either Syledis, local radio-transmitters in triangular configuration or a Decca positioning system. The former was used in applied surveys for raw materials, and is a positioning system matching the accuracy of DGPS, whereas the latter was used in scientific surveys, Poseidon cruises in the Skagerrak (Fig. 5) and Swedish cruises in the Kattegat Fält (1982). The Decca system provides positions with an accuracy of within 100–200 m, depending on the position relative to the radio-transmitters.

Gas depth contours

Putting together information from maps produced with the various reports on site surveys and raw materials plus maps from scientific surveys gave a fairly accurate overview of most gas occurrences within the Skagerrak-Baltic Sea region. A minimum depth of gas occurrence below seafloor was then assigned to each fix point on the shallow seismic lines within the gassy areas, using either ArcView or MapInfo GIS software. Contour lines were constructed using the drawing facilities in the above GIS software, and finally the contour lines were converted to ESRI® shape files using ArcView.

The minimum depth to gas was obtained from shallow seismic profiles from features indicating the presence of free gas. These features include (1) acoustic blanking, (2) acoustic curtains, (3) acoustic columns and (4) acoustic turbidity, examples of which have been presented by García-Gil et al. 2002). In addition, plumes of gas from older Pleistocene strata could be recognised in Holocene sediments above (Fig. 7). The following contour intervals were used for the Aarhus Bay area, offering the highest possible resolution: <0.5, 0.5–2, 2–4, >4 m. For the rest of the region, the resolution of the shallow seismic method enabled only two depth contours, 2–4 and >4 m, to be drawn. In most profiles, the top gas front was rather diffuse whereas in some cases the gas front appeared to be rather sharp, as can be seen in the examples from Aarhus Bay (Fig. 8).

Fig. 7
figure 7

Shallow seismic profile across the border area between plumes and shallow gas, ca. 6 km SE of Frederikshavn (FH; see Fig. 10)

Full size image
Fig. 8
figure 8

Shallow seismic (chirp) profiles in the Aarhus Bay area (for location, see Fig. 12). a Sharp upper gas front; b diffuse upper gas front

Full size image

Results and discussion

Based on existing shallow seismic profiles, a contour map indicating the sub-bottom depth of free gas in marine sediments was constructed. The gas contour map (Fig. 9) covers the Skagerrak-western Baltic Sea region and, in view of the relatively dense grid of shallow seismic lines (Figs. 5 and 6), it may be considered a complete map of all major shallow gas occurrences within that region (Fig. 9). It cannot be excluded, however, that minor gas occurrences in narrow depressions remain unidentified, as discussed below. For the Swedish sector of the southern Kattegat, which covers the muddy sediment area along the Swedish west coast (Fig. 1), no shallow seismic data were obtained. The presence of gas in the marine sediments within the contoured areas has been confirmed by extensive coring in (1) the Skagerrak-northern Kattegat area (van Weering 1982; Iversen and Jørgensen 1985; B.B. Jørgensen et al. 1990a; Dando et al. 1994b; Laier et al. 1996; METROL 2003) and (2) the western Baltic Sea area (Richardson and Davis 1998; METROL 2004).

Fig. 9
figure 9

Shallow gas depth-contour map

Full size image

The shallow gas map is stored digitally as an ESRI® shape file (Laier et al. 2005), which offers the possibility of viewing the map at any given scale using a suitable GIS software, as will be described below. Furthermore, details of the gas contour map may easily be compared with other relevant data, e.g. sediment type.

Occurrence and depth of free gas

As is visible even on a large scale (compare Figs. 1 and 9), almost all gas occurrences are found within areas of mud or sandy mud. This is also what one would normally expect since these sediments have higher organic matter contents compared to sandy sediments. However, exceptions exist, namely in the northern Kattegat where gas may be found also in sandy sediments (Fig. 9), mostly as gas plumes (Fig. 7). The depth of free gas below seabed shows remarkably little difference despite the relatively large variation in water depths within the region. In the Skagerrak–northern Kattegat area, free gas is present at all water depths from 20 m off the coast of Frederikshavn to approx. 360 m at the northern limit of the gassy area in the Skagerrak (Fig. 10). Yet, in most of the area, free gas is observed 2–4 m below seafloor. The northern limit of the gassy area in the Skagerrak almost parallels the 360 m water depth contour, although dissolved methane is likely to be present below seafloor at even greater water depths (Hass 1993; Werner, personal communication, 2004).

Fig. 10
figure 10

Shallow gas depth-contour map of the Skagerrak–northern Kattegat (for gas contour legend, see Fig. 9). Line FH-S indicates the location of the geological profile in Fig. 3. Line H-H′ corresponds to the shallow seismic profile in Fig. 11. D Seep location of Dando et al. (1994b), closed triangles methane-derived authigenic carbonates, open triangles seeps (Dando et al. 1994a; Zimmermann et al. 1997), closed circles (black/white) samples stations of Iversen and Jørgensen (1985) and Jørgensen (B.B. 1989a), open circles sample stations of Laier et al. (1996)

Full size image

Within the large gassy area, there are two minor areas which do not show shallow seismic evidence of free gas. In these two areas, the thickness of the Holocene marine sediments is less compared to adjacent areas due to the presence of ridges of glacial till, as is seen on the sparker profile across Hertha’s Flak (Fig. 11). Generally, the shallow seismic profiles across gassy to non-gassy areas indicate that the presence of free gas depends on both the thickness of Holocene sediments and water depth. For example, free gas is seen where sediment thickness exceeds 10 m at the western limit of the Kattegat shallow gas area, the water depths varying from 20 to 30 m (Larsen et al. 1986; METROL 2003). At the eastern limit, where water depths are 80–90 m, free gas appears only where the thickness of the Holocene exceeds 25–30 m (Fält 1982; METROL 2003). The thickness of Holocene deposits at the northern limit of the gas area in the Skagerrak is approx. 60–70 m (Dando et al. 1994a). It is fairly easy to understand why this is so. An increase in water depth (i.e. an increase in hydrostatic pressure) will increase the solubility of methane, meaning that more time is needed to build up the methane concentration necessary to exceed the solubility to form free gas. The increase in methane concentration versus time and sediment depth depends on the rate of methane formation and rate of diffusion of dissolved methane. If these were the only parameters controlling the presence of free gas, one might expect to see a greater variation in depth of free gas in the area. The fact the depth of free gas varies so little over the Skagerrak–northern Kattegat area indicates that, once free gas is formed, it will start to migrate upwards towards the seabed by some mechanism. As pointed out by Sills and Wheeler (1992), movement of gas bubbles due to buoyancy will not be possible in fine-grained sediments because of the high capillary forces in the pore throats. However, diffusion of gas in micro-cracks formed during sediment failure (Horseman et al. 1999; Boudreau et al. 2005) may explain movement of gas in fine sediments.

Fig. 11
figure 11

Seismic (sparker) profiles across Hertha’s Flak (for location, see line H-H′ in Fig. 10)

Full size image

Gas escape in the northern Kattegat

As mentioned above, the shallow gas GIS database facilitates comparison of gas occurrences with other relevant data from the region, for example, data on gas escape in the northern Kattegat area (as is shown in Fig. 10).The locations of sampling stations for previous studies on porewater chemistry concerning the sulphate-methane transition zone (Iversen and Jørgensen 1985; B.B. Jørgensen 1989a) as well as stations for radiocarbon dating of methane and organic matter (Laier et al. 1996) have also been indicated in Fig. 10. Finally, the location of the gas seepage area at the northern limit of the Skagerrak gassy area is shown (Dando et al. 1994a; Zimmermann et al. 1997). Integrating these data may improve our understanding of the controls of gas escape in the northern Kattegat, and may help us decide if and how seepage measurements from this area could be used in more general estimates of methane contributions to the atmosphere from the continental shelf.

No new data on methane seepage rates have been obtained since the rate measurements of Dando et al. (1994b) at the shoreline seepage location D in Fig. 10. However, new information concerning the geological controls of the gas seepage has been obtained, which will allow us to test the assumptions of Dando et al. (1994b) concerning the extrapolation of their seepage data to the whole of the Kattegat area. First facts about the seepage gas will be briefly mentioned; these are facts which were already known when Dando et al. (1994b) estimated the average gas escape from the Kattegat area to be approx. 2 × 103 g km−2 year−1. The gas is old microbial methane originating from Pleistocene deposits, as evidenced by stable isotopic and radiocarbon analysis (Laier et al. 1996). Minor isolated reservoirs in Pleistocene deposits, 70–120 m below surface, were known from the Frederikshavn area, 5 km north of the seepage location (Fig. 10) where gas had been exploited for household use during the 1930s and 1940s (Laier et al. 1992). Methane-derived authigenic carbonates associated with seeps (N.O. Jørgensen 1989b) are widespread in the northern Kattegat, and were known by local fishermen who reported most of the locations indicated in Fig. 10.

New radiocarbon dating results from other seeps in the Kattegat confirmed that the seepage gas is old. On the other hand, gas recovered from stations within the shallow gas area (PC125-PC130) proved to be much younger, 540–2,570 years b.p., showing that this gas had been generated locally (Laier et al. 1996). The youngest methane was collected close to the minor seeps at the northern limit of the Skagerrak, indicating that seeps were controlled by processes completely different from those in the Kattegat.

The stratigraphy of the almost 200-m-thick Pleistocene deposits (Fig. 3) was established by analysis of foraminifers from a number of exploration wells around the Fredrikshavn area (Knudsen 1985). Some wells showed a reversed succession of different zones in the Pleistocene, which Knudsen (1985) ascribed to poor quality of the drill samples. However, comparing with other data including shallow seismic profiles, Laier (2003) concluded that marine Pleistocene deposits of the area had been altered by glacial tectonics even at great depths. Thus, gas escaping from various seeps, including that examined by Dando et al. (1994b), is not necessarily derived from relatively deep gas accumulations, like those exploited previously and as was thought by Laier et al. (1992), but may originate from tectonized very shallow Pleistocene marine deposits. Shallow seismic profiles across other seepage areas, such as the Bubbling Reefs (Jensen et al. 1992) and Hertha’s Flak (Fig. 10), indicate that shallow Pleistocene deposits covered by sand is the prerequisite for gas seepage and the formation of methane-derived authigenic carbonate. Only in those areas of the Kattegat will methane escape from the seafloor. In none of the sampling stations within the contoured shallow gas was methane detected in porewater in the upper few decimetres of the seabed (Iversen and Jørgensen 1985; B.B. Jørgensen 1989a; Laier et al. 1996; METROL 2003), indicating a complete re-oxidation of methane at the sulphate-methane transition zone. Thus, we may conclude that the presence of gas seeps in the northern Kattegat is due to the effects of glacial tectonics on fairly local deposits of organic-rich Pleistocene marine deposits (Laier et al. 1992) in that area. Therefore, seepage measurements should not be averaged for the whole of the Kattegat area. We cannot yet propose a more realistic value for the total escape of methane from the Kattegat than that provided by Dando et al. (1994b). However, we can say that this estimate is valid for a much smaller area than assumed by Dando et al. (1994b), and that one should be cautious using such local measurements of seepage rates in general estimates of methane escape from the continental shelf.

Gas mapping as a monitoring tool

The Aarhus Bay area was chosen to test the possibility of using shallow seismic mapping as monitoring tool because the area is part of the nationwide monitoring program concerning the state of the aquatic environment (Kronvang et al. 1993; Christensen et al. 1998). Furthermore, several scientific studies (e.g. Tamdrup et al. 1994) have been performed on marine sediments in the bay. Shallow seismic surveys for environmental purposes have not been performed previously in the area. Therefore, the area most vulnerable with respect to the escape of methane and toxic hydrogen sulphide, i.e. the area having the shallowest gas occurrence, had not been identified yet. One shallow seismic survey for raw materials performed in 1983 (Fig. 6) was used to identify areas for further investigations of gas occurrences, including new shallow seismic surveys and sediment coring (Jensen and Laier 2003). Based on existing as well as new shallow seismic data, a shallow gas depth-contour map was constructed (Fig. 12). Although most of the seabed in the bay consists of mud (Fig. 2), the distribution of shallow gas is variable (Fig. 12). This is due to fact that Holocene sediments vary in thickness and composition, as is seen in the geological profile across the bay in Fig. 4. The different units reflect different depositional environments due to change in relative sea level during the Holocene.

Fig. 12
figure 12

Shallow gas depth-contour map of the Aarhus Bay area. Lines A and B indicate locations of shallow seismic profiles in Fig. 8a and b, respectively. Circles indicate sampling stations for porewater analysis

Full size image

The theory behind the idea of using shallow seismic mapping for monitoring the vulnerability of the bay in case of long periods of oxygen deficiency is based on the assumption that the depth of free gas is somehow related to the depth of the sulphate-methane transition zone. Periods of oxygen deficiency will occur and have occurred after algae blooms due to eutrophication. Higher inputs of organic matter will consume oxygen in the bottom waters and may move the sulphate-methane transition zone closer to the seabed. In seabed sediments, organic matter mineralisation occurs not only via sulphate reduction but also via the reduction of iron(III), which may act as buffer. In case of long-term eutrophication, this buffer may be used as is indicated by the studies of Rysgaard et al. (2001) and Jensen et al. (2003). If a simple relationship between shallow gas depth and sulphate-methane transition zone can be established by shallow seismic mapping, then this may help to document long-term effects concerning the vulnerability of the Aarhus Bay in case of extended periods of oxygen deficiency. Preliminary data on porewater chemistry across the gassy-non-gassy area (Fig. 13) indicate that such a relationship may be established, although more data from a wider area are needed.

Fig. 13
figure 13

Preliminary results of porewater analysis (Fossing, personal communication). M2–M5 stations from north to south are indicated in Fig. 12. Vertical lines indicate error bars on methane saturation points

Full size image

Viewing shallow map details using GIS software

A CD-ROM containing the electronic version of the shallow gas depth-contour map has been produced and may be obtained from the authors. Opening the proper files using ArcExplorer, freely available from ESRI, one will be able to view different areas within the Skagerrak-western Baltic Sea region. Not only shallow gas data may be viewed but also other relevant data can be seen simply by putting tick marks on the list shown on the left-hand side of the window. A short introduction is also included which may be viewed using Microsoft PowerPoint (Laier et al. 2005).

Conclusions

The construction of a digital version of a shallow gas depth-contour map based on a fairly dense grid of existing shallow seismic profiles facilitates comparison with other data, allowing a number of conclusions to be drawn.

  1. 1.

    Free gas usually occurs only in mud and sandy mud but is seen only when the sediment exceeds a certain thickness, depending on water depth.

  2. 2.

    In spite of the sediment thickness-water depth dependence, gas depths below seafloor vary only little within the region, regardless of water depth.

  3. 3.

    It is therefore concluded that free gas, once formed, moves relatively quickly towards the seafloor, compared to the diffusion rate of dissolved methane.

  4. 4.

    Gas seeps in sandy sediments are confined to a restricted area of the northern Kattegat characterised by shallow organic-rich Pleistocene marine deposits.

  5. 5.

    Therefore, it is not safe to extrapolate seepage rates from this area to make general estimates on methane escape to the atmosphere from the continental shelf.

  6. 6.

    In open marine areas, which do not suffer from oxygen deficiency, all methane is being re-oxidised within the sediments and no escape of methane occurs from the seabed.

  7. 7.

    In near-coastal areas suffering from oxygen deficiency due to eutrophication, shallow seismic mapping may be used to monitor the vulnerability of the area.