We analysed the long-term variations in grain-size distribution in
sediments from Gåsfjärden, a fjord-like inlet in the southwestern
Baltic Sea, and explored potential drivers of the recorded changes in the
sediment grain-size data. Over the last 5.4 thousand years (ky) in the study
region, the relative sea level decreased 17 m, which was caused by isostatic
land uplift. As a consequence, Gåsfjärden was transformed from an
open coastal setting to a semi-closed inlet surrounded by numerous small
islands on the seaward side. To quantitatively estimate the morphological
changes in Gåsfjärden over the investigated time period and to
further link the changes to the grain-size distribution data, a digital
elevation model (DEM)-based openness index was calculated. The largest
values of the openness indices were found between 5.4 and 4.4 cal ka BP,
which indicates relatively high bottom water energy. During the same period,
the highest sand content (
Sedimentary grain-size distribution provides important information regarding depositional conditions and has been widely analysed in both modern samples and sediment cores (e.g. Tanner, 1992; Yang et al., 2008; Virtasalo et al., 2014). Grain-size distribution is generally governed by sediment inputs and hydrodynamic energy conditions. The higher the energy conditions, the larger proportion of coarse grains (Dearing, 1997; Jönsson et al., 2005). Water depth, wind direction and strength, basin morphometry, and man-made constructions such as dams could influence bottom water hydrodynamics and may lead to different characteristics in grain-size distributions. The Baltic Sea is connected with the North Sea through the narrow Danish Straits. Although tidal activity can strongly influence grain-size distribution in coastal regions, as shown by Zhang et al. (2002), the tidal amplitude recorded in the Baltic Sea is only a few centimetres (Ekman and Stigebrandt, 1990) and therefore its impact on sediment grain size is not considered in this region. Instead, wind conditions and coastal morphometry are considered to be the most important factors that influence the sedimentary grain-size distribution in the Baltic Sea coastal zone (Lehmann et al., 2002; Jönsson et al., 2005; Al-Hamdani and Reker, 2007).
During the Holocene, the Baltic Sea has experienced several stages modulated by global sea-level changes as well as isostatic land uplift (Björck, 1995; Andrén et al., 2011). As the late Weichselian ice sheet retreated, land uplift during the deglaciation and the Holocene resulted in shoreline displacements in the coastal zones of the Baltic Sea. A maximum of 60 m decline in relative sea level (RSL) has been recorded over the last 6 thousand years (ky) (Påsse and Andersson, 2005), leading to basin isolations and long-term changes in the coastal morphometry (Eronen et al., 2001). These changes in the coastal morphometry variations may potentially be linked with variations in grain-size distributions, as shown from a coastal inlet in the southwest of Baltic Sea (Ning et al., 2016a).
To examine the impact of coastal morphometry changes on grain-size distribution in a long-term perspective, quantifying the morphological changes, e.g. water depth and cross-section areas, could be useful. Achieving this quantification is data-demanding, since it requires (1) high-resolution bathymetry data, (2) the correct cross-section area and (3) sedimentation rates. However, all these data are difficult to obtain and make the quantification difficult. Alternatively, through using digital elevation maps, Lindgren (2011) proposed a geographical information system (GIS)-based wave fetch index, named filter factor, to estimate coastal morphometry. The result of the quantified coastal morphometry was found to be significantly correlated with bottom water dynamics (Persson and Håkanson, 1995) and deep-water turnover time (Persson and Håkanson, 1996). There are also other GIS-based indices that exist for describing coastal openness and wave exposure (Ekebom et al., 2003; Tolvanen and Suominen, 2005), and these GIS-based methods have also been applied to investigate sediments grain-size distributions from lakes and coastal zones (Håkanson, 1977; Lindgren and Karlsson, 2011). However, these aforementioned indices are restricted to depicting the modern coastal morphometry and have not yet been employed in a palaeoenvironmental context.
In this study, we proposed an openness index using digital elevation model (DEM) data, and this approach could provide an opportunity to estimate long-term coastal morphometry variations for the Holocene. Furthermore, we innovatively used the coastal openness index for grain-size data interpretations. The aim of the study is to (1) present a method for quantifying openness changes in coastal region that experienced large relative sea level as well as shoreline changes and to (2) link the estimated openness index with the long-term sediment grain-size distribution.
Changes of relative sea level in the study area over the last 5.4 ky based on the empirical model by Påsse and Andersson (2005).
Gåsfjärden is a semi-enclosed fjord-like inlet located on the
southeast Swedish Baltic Sea coast (Fig. 1a). It has a restricted water
exchange with the open Baltic Sea through a narrow and shallow strait
(
The light detection and ranging (lidar)-based DEM data of the study region
(Fig. 1c) were obtained from the Swedish mapping agency, Lantmäteriet
(
Sediment cores were collected at station VG31 (57
Illustrations showing the variations in seaward radial lines
intercepted with land over the last 5.4 ky:
The calculation of the openness index (Fig. 3) has been modified on the basis
of the method described by Lindgren (2011) and the fetch-length method of
Ekebom et al. (2003). The following steps for estimating openness index
variations were taken in ArcGIS 10.3:
The coring site was identified in the DEM. Using the coring site as a starting point, two sets of 180 The RSL changes with a 100-year interval were applied to the present-day
DEM. For every 100 years, a new DEM was generated and the RSL changes were
based on the age–RSL relationship in Fig. 2. For every 100 years, the grid cells in the generated DEM were classified as
sea or land based on the elevation. The raster DEMs were converted to land and sea polygons for vector
calculation in ArcGIS, and the radiating lines generated in step 2 were
divided into smaller segments when the lines were intersected by the land
polygons. The lines originating from the coring site and that came into contact with
the land were selected (see Figs. 4 and 5). The seaward and landward openness
indices were calculated as the average length of the selected radiating
lines.
Illustrations showing the variations in landward radial lines
intercepted with land over the last 5.4 ky:
Calculated landward
Calculated landward
The calculated openness indices using 1
The openness indices with different intervals have been estimated in order to
determine an optimal interval for applying this index for the study area. The
seaward and landward openness indices, calculated with 15 and 10
Correlation (
The openness indices calculated with a 1
Erosion from the surrounding islands since 5.4 ka has most likely occurred, but it has been rather limited as these islands are mostly rocky with little soil cover. It may, however, result in a flux of relatively coarser grains to the coring site during uplifting of the land. As the uplift process has shown to be linear (see Fig. 2), we might expect to see a rather linear change in the grain-size data if the uplifting of the land had played the dominant role. However, the sand content and silt / clay ratios exhibit strong year-to-year variations, which indicates other factors than land uplifting could also participate in influencing grain-size distribution. For instance, coarse grains, such as sand, can also be transported to the coring site through storm events and intense wave action, sea ice or drifting seaweed. However, their impacts are not explicitly included in the openness indices. Furthermore, the recorded large variability in the sand content within the last millennium may be linked with catchment disturbance from human activities (Karlsson et al., 2015; Ning et al., 2016a).
The silt / clay ratios also reflected bottom water energy, with higher values
indicating higher energy conditions. The silt / clay ratio was
Our DEM-based calculations of coastal openness indices have shown to be a useful tool when interpreting long-term sedimentary grain-size data. A relatively high RSL was linked with large coastal openness and higher hydrodynamic energy, which in turn was well reflected in the seaward openness index. The higher values of both sand content and the seaward openness index were recorded in the early part of the record, indicating that coastal morphology (presented by openness index) strongly influenced sand distribution. The differences in temporal dynamics of sand content and silt / clay ratios indicate that different grain-size sediments responds differently to the changes in hydrodynamic energy. The significant decline in silt / clay ratios between 4 and 0.1 ka demonstrated that the long-term impacts of coastal openness on the finer grain-size sediment distributions. Our DEM-based openness index can be easily applied to other coastal settings that have experienced large sea-level changes over time. The index could also be further used in predicting future dynamics by combing information about sea-level changes in a warmer future.
All data used in this manuscript are available at Pangaea
(
Z. Wu is thanked for help with data analysis. The project was funded by
FORMAS Strong Research Environment: Managing Multiple Stressors in the Baltic
Sea (217-2010-126). We thank the captain and crew of R/V