Edinburgh Explorer Postglacial relative sea-level changes in northwest Iceland: Evidence from isolation basins, 1 coastal lowlands and raised shorelines

Relative sea-level (RSL) data provide constraints on land uplift associated with former ice 13 loading and can be used to differentiate between contrasting ice unloading scenarios. Isolation 14 basin, coastal lowland and geomorphological evidence is employed to reconstruct RSL changes in 15 northwest (NW) Iceland, which may have experienced contrasting uplift patterns. Under local 16 (NW) uplift, highest RSL would be expected in central Vestfirðir, whereas highest RSL would be 17 closest to the main ice-loading centre under regional (central Iceland) uplift. Four new RSL 18 records are presented based on 16 sea-level index points and 4 limiting ages from sites principally 19 focussed along a transect away from central Iceland. The new RSL records highlight spatial 20 variability of Holocene RSL changes and provide constraints on deglaciation. There is an increase 21 in marine limit elevation with proximity to the proposed principal ice loading centre in central 22 Iceland. Highest recorded marine limit shorelines are found in Hrútafjörður-Heggstaðanes 23 (southeast), the lowest in Hlöðuvík and Rekavík bak Látrum (north), and at an intermediate 24 elevation in Reykjanes-Laugardalur (central Vestfirðir). Evidence from Breiðavik-Látrar records 25 early rapid deglaciation in Breiðafjörður or a complex interplay of multiple uplift centres. RSL fell 26 rapidly following deglaciation in several locations as a result of the quick response of the Icelandic 27 lithosphere to unloading. The RSL data along the transect show an uplift pattern consistent with 28 extensive regional glaciation emanating from central Iceland, which could have implications for ice 29 sheet configuration and patterns of deglaciation, glacio-isostatic adjustment modelling and the 30 volume of meltwater input into the North Atlantic. 31

lithosphere to unloading. The RSL data along the transect show an uplift pattern consistent with 28 extensive regional glaciation emanating from central Iceland, which could have implications for ice 29 the potential to produce high-resolution data to identify the location and thickness of former ice 48 loading through constraint of the marine limit, the establishment of deglacial timing and the 49 patterns of Lateglacial to Holocene relative sea-level changes. In turn, these data act as important 50 constraints for glacio-isostatic adjustment (GIA) models, which can further assist in the testing of 51 ice loading hypotheses, lithospheric and mantle viscosity characteristics. This paper provides new 52 relative sea level (RSL) data from northwest (NW) Iceland, which reflect post-(de)glacial loading 53 and unloading of the crust as a result of near-equilibrium glacio-isostatic conditions during 54 deglaciation (Norðdahl and Ingólfsson 2015). Establishing the lateral and vertical extents of the 55 LGM IIS, associated ice volumes and patterns of deglaciation, is crucial, due to Iceland's location 56 close to sensitive areas of deepwater formation in the Nordic Seas and northern North Atlantic 57 across an impervious rock sill that controls tidal inundation (e.g. Evans, 2002, Long et al, 65 2011). A series of stages of basin isolation have been identified (e.g. Lloyd and Evans, 2002) and 66 analysis of sediment and microfossil datasets allows the identification of three isolation contacts -67 diatomological, hydrological and sedimentological -which can subsequently be linked to positions 68 within the tidal frame (Kjemperud, 1986). Radiocarbon dates at these isolation contacts provide 69 constraints on the timing of RSL change and the resulting RSL curves may in turn determine 70 patterns of postglacial land-level change (e.g. Long et al., 2011), allowing an assessment of former 71 ice loading patterns. 72 Coastal lowlands are situated close to present sea-level and encompass the environment from 73 mud flat to above high marsh conditions, and have the potential to record past RSL changes where 74 sufficient accommodation space is available to record changes in environmental conditions. 75 Coastal lowland environments may therefore encompass saltmarshes, which have previously been 76 used in Iceland to reconstruct patterns of RSL change (e.g. Gehrels et al., 2006;Saher et al., 77 2015).   The sites were chosen to exploit RSL changes across a potential former ice loading centre above 192 Vestfirðir. The sites in Hlöðuvík and Rekavík bak Látrum, Reykjanes-Laugardalur and  Heggstaðanes form the research transect, with secondary sites at Breiðavík and Látrar at the 194 mouth of Breiðafjörður. 195 Figure 3 illustrates the hypothesised patterns of RSL change at the three main sites along the 196 transect under two glacio-isostatic uplift scenarios. Under a regional uplift scenario, with ice 197 loading emanating from central Iceland (black line, Fig. 3), it is proposed that the marine limit 198 elevation would increase with proximity to the former centre of unloading (and therefore uplift) in 199 central Iceland. Alternatively, the highest marine limit elevation would be found in  Laugardalur under a local uplift scenario (grey dashed line, Fig. 3

RSL reconstruction 211
In order to assess patterns of RSL change in NW Iceland, we collected isolation basin and coastal 212 lowland sediment samples from the marine limit to present sea level in each of the four field 213 locations. The isolation basins were selected using the criteria outlined by Long et al. (2011), 214 ensuring a suitable size (<1 km 2 ), depth (<10 m) and spacing of sites in each research location 215 sediments, a grid of cores established the lowest high point in the underlying bedrock. The 217 elevation of the isolation basin sill was measured relative to mean high water spring tide (MHWST) 218 in the field using an Electronic Distance Meter (EDM; ± 0.1 m) and subsequently corrected to mean 219 sea level (MSL or m a.s.l.) using tide tables for the nearest tide station (Admiralty Tide Tables,  220   2006). 221 The stratigraphy of each isolation basin was established by coring perpendicular transects with a 222 gouge corer from infilled sections or from the rear of a boat when a lake was present. Samples 223 extracted were described using the Troels-Smith (1955) classification scheme. Following initial 224 survey, a core was extracted from the deepest point along the transect using a Russian corer 225 (Jowsey, 1966). 226 Diatom preparation followed the standard procedures outlined by Palmer and Abbott (1986) and 227 diatoms were classified using a range of sources (e.g. Brun, 1965;Foged, 1974;Hartley, 1996).  In order to establish the elevation of former RSL, a correction is required based on the indicative 251 meaning of the dated point in the diatom assemblage (see Shennan et al., 2015), with the error 252 comprised of elevation uncertainty, sill determination uncertainty and the indicative range of the 253 assemblage. Establishment of a series of SLIPs for each research location has allowed the 254 production of a series of new RSL curves for NW Iceland. Full details of site stratigraphies, diatom 255 assemblages and tephra geochemical results are presented as Supplementary Information. 256

Results -New sea-level index points for NW Iceland 257
Results of stratigraphic, diatom and chronological analyses are divided into the four principal 258 geographical locations investigated as part of this research (Fig. 2). In total, 16 new SLIPs and 4 259 limiting points have been generated for NW Iceland. 260 A and B (Fig 2); five sites) 261

Hlöðuvík and Rekavík bak Látrum (Area
Four lake basins and one raised shoreline were surveyed in Rekavík bak Látrum and Hlöðuvík 262 ( Fig. 2A and B). The majority of these sites are found in Hlöðuvík, where the marine limit can be 263 HD3 is the lowest basin sampled in Hlöðuvík (Fig. 2B). The stratigraphy is characterised by a 270 basal tephra deposit overlain by olive green limus, sandy gravel and uppermost turfa peat, with a 271 visible tephra at the base of the analysed core (Fig. 4). Geochemical analysis of a dark grey/black 272 tephra deposit allows identification as being part of the Saksunarvatn sequence of tephras (Fig. 5). 273 It is now apparent that there were multiple eruptions of Grímsvötn between 9.9 cal ka BP and 10.4 that the 'Saksunarvatn' tephra found in our cores correlates with this too, but again cannot be 282 confirmed. Diatom analysis shows freshwater conditions dominate at the site, suggesting that RSL 283 was lower than the sill elevation at 10.2 cal. ka BP. This also acts as a limiting age for marine limit 284 formation. HD2 is a small basin situated west of HD3 (Fig. 2B). The site stratigraphy is comprised of a basal 287 silt, overlain by olive-green limus and turfa peats. No tephra deposits were found at the site. The diatom assemblage for HD2 is dominated by freshwater conditions (Fig. 4) suggesting that the site 289  HD1 is a small but the uppermost basin in Hlöðuvík (Fig. 2B). The site stratigraphy is 299 characterised by a basal blue-grey clay overlain by olive-green limus. Above this layer, an organic 300 rich layer is evident, covered by turfa peat. The diatom assemblage is dominated by freshwater 301 taxa (Fig. 4) and therefore the site acts as a limiting elevation for former RSL. The raised shoreline in Hlöðuvík was surveyed using an EDM in order to establish the highest 305 postglacial RSL in the region (Fig. 2B). Surveying established the minimum elevation of the 306 marine limit at ~18 m a.s.l., just above the value reported by Hjort et al. (1985, 10-15 m a.s.l.). The 307 raised beach is characterised by a lower till deposit, overlain by marine sediments, with a 308 subsequent upper till. There was a lack of dateable material within the proposed 'marine' 309 sediments, meaning that it has not been possible to establish an accurate chronology for this 310 feature. However, the discovery of the Saksunarvatn tephra (10.2 cal. ka BP) in the basins above 311 the marine limit acts as a limiting age for feature formation.  REK1 is situated aside a kettle-hole lake (Hálsavatn) in Rekavík bak Látrum and represents the 319 westernmost field site in the region ( Fig. 2A). The site stratigraphy is comprised of a basal tephra-320 rich gravel overlain by clay rich silts, silty limus and an uppermost organic rich limus (Fig. 4). The 321 Saksunarvatn tephra was identified at the base of the core following geochemical analysis of a 322 black tephra deposit (10.2 cal. ka BP; Fig. 5). There is a clear dominance of freshwater conditions 323 at the site, but the occurrence of limited numbers of brackish taxa suggests occasional inundation 324 by highest times or storm events at the base of the core. A radiocarbon sample at 330 cm 325 produced an age of isolation of 9130 -9412 (9.2 k) cal. a BP (Table 1), which provides a limiting 326 (minimum) age for marine limit formation in the region. 5.2 Reykjanes-Laugardalur (Area C (Fig. 2); 10 sites) 342 Ten sites in the Reykjanes-Laugardalur area were investigated from the marine limit to present sea 343 level. Reykjanes-Laugardalur represents the centre-point for the research transect to evaluate 344 contrasting glacio-isostatic uplift scenarios due to different uplift centres (Fig. 3). BB1 is a small embayment close to the farm at Vatnsfjörður, (Fig. 2C). A sample comprising basal 347 gravel rich silt, overlain by a gravel rich turfa peat and uppermost gravel rich silt layer was collected 348 close to the present beach (Fig. 7). Diatom analysis reveals marine dominance in the lowermost 349 unit, with a limited (~20%) freshwater component to the diatom assemblage. Within the turfa peat, 350 freshwater influence increases with a minor brackish component. Diatom preservation in the 351 uppermost sediment unit was poor and provided insufficient numbers for a reliable sample for 352 analysis. A radiocarbon date from the turfa peat (bulk sample) returned a 'modern' age for the 353 deposit (Table 1). RK6 is a small basin, found north of the present airfield on Reykjanes (Fig. 2C). The sediment 370 profile at RK6 comprises a basal olive green mixed organic material, overlain by a lower peat layer, 371 olive green humified organic material and middle peat layer. Above this, is an upper olive green 372 organic layer, overlain by an upper turfa peat layer. The diatomological isolation contact is 373 identified at 100 cm (Fig. 7) shown by a reduction in brackish conditions at the site. A bulk organic 374 radiocarbon sample at 100 cm produced an age for the SLIP of 9139 -9432 (9.3 k) cal. a BP 375 (Table 1). VHF1 is situated close to an archaeological site and present farm at Vatnsfjörður (Fig. 2C). The 379 stratigraphy is made up of a basal blue-grey sandy clay, overlain by extensive peats. A number of 380 sediment samples were extracted through the profile. Diatom analysis shows the core dominated 381 by freshwater conditions, with a weak brackish signal at the base (Fig. 7). A radiocarbon sample at 382 69 cm provides a marine limiting age of 5584 -5711 (5.6 k) cal. a BP (Table 1) RK3 is situated between RK6 and RK10, south of the present airfield on Reykjanes (Fig. 2C). A 392 sediment sample comprised a basal brown silty mixed organic material, overlain by a grey silt, 393 brown mixed organic material and upper peat layer. The diatom assemblage shows a transitional 394 sequence and can be divided into five zones. A radiocarbon sample at 147 cm produced a timing 395 of isolation of 3829 -4071 (3.9 k) cal. a BP (Fig. 7; Table 1). There is a clear reduction in marine 396 influence at the site over the course of the diatom record (Fig. 7). RK10 is a predominantly infilled basin on the Reykjanes peninsula, situated between RK3 and the 399 airfield (Fig. 2C). The site stratigraphy is characterised by a basal gravel, extensive limus deposits 400 and turfa peat. Diatom analysis highlights two distinct zones (Fig. 7). The diatomological isolation 401 contact is clearly evident at 237 cm and a bulk radiocarbon sample at 238 cm returned an age for 402 the SLIP of 9798 -10190 (10.0 k) cal. a BP (Table 1). In addition, the Saksunarvatn tephra was 403 identified by geochemical analysis at 248 cm, providing a second (minimum) age of 10.2 cal. ka BP 404 (Fig. 5). There is a clear reduction in marine influence at the site, suggesting a RSL fall at the 405 location. shortly after the transition from silt to limus. Geochemical analysis of these dark grey deposits has 412 identified the Saksunarvatn tephra at 163 cm (10.2 cal. ka BP; Fig. 5). The diatom assemblage 413 from VAT1 can be divided into four distinct zones (Fig. 7). The diatomological isolation contact is 414 identified at 204 cm from which a radiocarbon sample produces an age of 9918 -10216 (10.1 k) 415 cal a BP (Table 1)  GR1 is a large basin situated close to the local marine limit at ~25 m a.s.l. in Laugardalur (Fig. 2C). 419 The core from the northern section of the present lake basin is characterised by a basal silty brown 420 limus overlain by an olive green limus layer. The diatomological isolation contact is identified at 421 212 cm, with a radiocarbon sample generating an age of 10444 -10724 (10.6 k) cal. a BP for the 422 SLIP indicating a reduction in marine influence of a brackish environment at the location (Fig. 7). VAT2 is a large basin found above and south of site VAT1 on the Vatnsfjarðarnes peninsula (Fig.  426 2C). The collected core was comprised of a basal gravel, overlain by silty clay, limus and peat 427 layers. The diatom assemblage is dominated by freshwater conditions. There is a short-lived 428 brackish component at 428 cm, which may represent a storm event or brief marine incursion of the 429 basin. A clear transitional sequence is not evident at the site, suggesting that the site was situated 430 above the influence of marine conditions. A radiocarbon sample at 428 cm produced an age of 431 11712 -12067 (11.9 k) cal. a BP (Table 1) for a short-lived brackish episode. 432

Laugardalur (LG1) 433
The raised shoreline at Laugardalur provides an elevation for the local marine limit in Reykjanes-434 Laugardalur (Fig. 2C). The distinctive feature is found at the mouth of Laugardalur valley and was 435 surveyed in a number of locations using an EDM to ~25/30 m a.s.l. A lack of dateable material 436 means that it has not been possible to directly date the feature, although diatom analysis from GR1 437 provides a limiting age for formation of the shoreline. 438

RSL curve for Reykjanes-Laugardalur 439
Eight new SLIPs and the surveyed marine limit produce a new RSL curve for the region (Fig. 6; 440 Table 1). The new RSL curve for Reykjanes-Laugardalur shows RSL rapidly falling from the 441 marine limit at ~25/30 m a.s.l., which may have been formed at ca. 10.6 -11.9 cal. ka BP, to below 442 present sea level by ca. 9.3 cal. ka BP. RSL then rose above present sea level to a mid-Holocene 443 highstand at ~4 m a.s.l. between ca. 4 and 5.8 cal. ka BP, before falling to present sea level (Fig.  444   6). The regression sequences from RK6 and VHF1 mean that sea-level must have risen to or close 445 to the elevation of these sites during the early to mid-Holocene. 446 (Fig. 2); eight sites) 447

Hrútafjörður-Heggstaðanes (Area D
Eight isolation basin and coastal lowland sites were investigated in the Hrútafjörður-Heggstaðanes 448 area, which represents the innermost research location along the principal research transect 449 through NW Iceland (Fig. 3). KB4 is found to the southwest of KB2 on the Kolbeinsárnes peninsula (Fig. 2D). The sediment core 467 contains a lower silty clay, organic-rich limus and overlying turfa peat (Fig. 8). The diatom 468 assemblage shows a transition from brackish-marine to freshwater dominance (Fig. 8). KB1 is a small basin situated north of KB2 and northeast of KB4 (Fig. 2D). The stratigraphy 475 comprises of a basal blue-grey clay with silt, organic rich silt, olive-green limus with abundant 476 rootlets and a distinct uppermost olive-green limus layer (Fig. 8). The diatom assemblage 477 represents a gradual decrease in marine influence at the site (Fig. 8). A bulk sediment sample for 478 radiocarbon analysis from the diatomological isolation contact at 65 cm returned an age of 2185 -479 2465 (2.3 k) cal. a BP for RSL falling below the SLIP. SN1 is situated north of SN2 on the Heggstaðanes peninsula (Fig. 2D). A transect of 4 cores 482 produced a stratigraphy comprising a basal blue-grey silt, overlying limus layer and surface turfa 483 peat deposits. A tephra deposit was evident between the silt and limus layer at 578 cm, identified 484 as the Saksunarvatn tephra (10.2 cal ka BP) following geochemical analysis. Diatom analysis 485 reveals a transitional sequence from brackish to freshwater dominance (Fig. 8). A radiocarbon 486 sample at the apparent diatomological isolation contact at 610 cm produced a minimum age of 487 10814 -11216 (11.1 k) cal. a BP for the SLIP. SN2 is a large infilled basin also situated on the eastern side of the Heggstaðanes peninsula (Fig.  490   2D). The site stratigraphy was established through 4 cores and is characterised as a basal silt 491 overlain by organic rich limus containing a distinct tephra layer and an uppermost turfa peat layer. 492 The Saksunarvatn tephra was identified within the limus deposit at 509 cm, providing an age of 493 10.2 cal ka BP. In addition, a radiocarbon sample from an apparent diatomological isolation 494 contact at 610 cm (Fig. 8)  stratigraphy was established by a transect of 3 cores and comprises a basal gravel with overlying 500 blue-grey silts and clays, mixed organic sediments and uppermost peat layer. A number of 501 individual tephra layers were identified within the sedimentary profile. The diatom assemblage can 502 be divided into three distinct zones (Fig. 8). A radiocarbon sample was analysed from 612 cm and 503 returned an age of 11191 -11311 (11.2 k) cal. a BP for the SLIP shown by the transition from 504 marine, brackish to freshwater dominance in the diatom flora. The Saksunarvatn tephra was also 505 identified at 592 cm, providing additional chronological control for the site (10.2 cal ka BP). AH2 is a small basin ~110 m west of AH1 in innermost Heggstaðanes (Fig. 2D). A basal blue-grey 508 sand, overlain by silty clay and an olive-green limus is present within the cores. A dark grey tephra 509 was identified as Saksunarvatn at 594 cm following geochemical analysis, providing an age of 10.2 510 cal ka BP ( Fig. 5 and 8). Diatom samples were analysed throughout the core sample showing 511 freshwater dominance but the lowermost samples provided insufficient diatoms to ensure a reliable 512 count (Fig. 8). A radiocarbon sample from 632 cm provided a limiting age for the deposition of 513 organic material and the site was therefore above RSL at 11109 -11242 (11.2 k) cal. a BP. AH1 is the highest basin investigated in Hrútafjörður-Heggstaðanes, situated to the northeast of 516 AH2 (Fig. 2D). The sediment stratigraphy was established through a transect of 3 cores and 517 comprises basal blue-grey clay and silty clay overlain by an olive-green limus. A dark grey tephra 518 layer was evident at 609 -612 cm, which was identified as the Saksunarvatn tephra following 519 geochemical analysis, providing an age of 10.2 cal ka BP ( Fig. 5 and 8). In total, nine diatom 520 samples were analysed, although the lowermost samples failed to produce sufficient diatoms to 521 ensure a valid count (Fig. 8). A radiocarbon sample at 613 cm produced an age of 10781 -11174 522 (11.0 k) cal. a BP and acts as a limiting age for the site, although this should be treated with some 523 caution, given the close proximity to the Saksunarvatn tephra (10.2 cal. ka BP). 524

RSL curve for Hrútafjörður-Heggstaðanes 525
Six new SLIPs and two limiting points have been produced in Hrútafjörður-Heggstaðanes, the 526 innermost location along the research transect in NW Iceland. These new SLIPs have allowed the 527 construction of a tentative new RSL curve for the region, highlighting initial RSL fall and more 528 recent RSL changes (Fig. 6). The lack of mid-elevation sites in the region means that it has not 529 been possible to constrain RSL changes between ca. 11200 and 2400 cal. a BP (Fig. 6). act as limiting RSL points, suggesting that RSL has most likely been below this level since 534 deglaciation. 535 (Fig. 2); one site) 536

Breiðavík-Látrar (Area E
A series of sites were investigated in the Breiðavík-Látrar area in order to investigate the high 537 marine limit elevations recorded in the region. A number of higher elevation sites recorded 538 evidence for marine influence, yet suffered from poor chronological control, and thus are not 539 presented here. Consequently, only one site is presented here in full, from close to present sea 540 level (Fig. 9). The elevational data from the higher sampled sites can however be seen in Fig. 6. The BR10 locality is situated in Breiðavík, a large bay on the westernmost part of Iceland (Fig. 2E). 544 The site stratigraphy can be summarised as a basal sand overlain by silt-rich limus and organic 545 rich silts, with visible shell remains likely deposited into the basin by aeolian transport. The diatom 546 record shows a reduction and subsequent increase in marine influence at the location, with the 547 diatomological isolation contact therefore identified at 218 cm (Fig. 9). A bulk radiocarbon sample 548 from 218 cm produced an age of 1301 -1407 (1.4 k) cal. a BP. 549 One new SLIP has been generated for the region (see Table 1), which is supported by the local 555 marine limit at ~85 m a.s.l. by Norðdahl and Pétursson (2005 (Pétursson, pers. comm, 2016). Additional sites with brackish diatoms were recorded at ~64 m, 67 558 m and 73 m a.s.l. (Fig. 9) but suffered from poor chronological control. It is likely that organic 559 productivity was low immediately following deglaciation and as a result, limited organic material 560 was available for dating within these sediment sequences. The ages generated are not consistent 561 with the elevation of the samples and thus probable timing of isolation. In Hlöðuvík and Rekavík bak Látrum, the new RSL data provide constraint on marine limit 573 formation. The marine limit in Hlöðuvík is recorded at ~17.5 ± 1.0 m a.s.l. and is characterised by 574 a basal till overlain by marine sediments and an upper till (15-26 m a.s.l., Hjort et al., 1985). Lake 575 basin samples from close to the marine limit demonstrate entirely freshwater assemblages and 576 therefore support the interpretation of this feature. The presence of the Saksunarvatn tephra in 577 these lake basin sediments suggests that the lower part of the valley was ice free by 10.2 cal ka 578 BP and provides a limiting (minimum) age for local marine limit formation. This interpretation is 579 supported by Hjort et al (1985) who identified that the Saksunarvatn tephra was deposited close to 580 sea-level, suggesting that RSL was below the elevation of the lake basin sites at 10.2 cal. a BP. 581 In Rekavík bak Látrum, following deglaciation RSL fell from the local marine limit at 15-25 m a.s.l. 582 (Hjort et al., 1985). This RSL fall is constrained by the new RSL data to between 9.3 cal. ka BP 583 and 10.2 cal. a BP, which acts as both the minimum age for marine limit formation and 584 deglaciation. Additional sites explored in Rekavík bak Látrum demonstrate extensive gravel and 585 sand deposits, limiting the potential for reconstructing environmental change for the westernmost 586 section of the study area. 587

Reykjanes-Laugardalur 588
Following deglaciation, RSL fell from the local marine limit at ~30 m a.s.l. at ca. 10.6 cal ka BP until 589 9.3 cal. ka BP, after which RSL fell below present (Fig. 6). The RSL record generated in 590 Reykjanes-Laugardalur therefore demonstrates that deglaciation may have occurred at a later date 591 (10.6/11.9 cal. ka BP, GR1/VAT2) in Ísafjarðardjúp than north of Breiðafjörður (Lloyd et   The difference in deglacial age is possibly the consequence of later glacier retreat at the innermost 605 part of the fjord system due to lesser ingress of the warm Irminger Current, which would have 606 become fully established in NW Iceland by ca. 10.2 cal. ka BP (Ólafsdóttir et al., 2010). 607 Following the RSL fall below present sea level in the early Holocene (Fig. 6), a transgression must 608 have occurred in the mid-to early Holocene, with the Reykjanes-Laugardalur RSL curve showing 609 an associated regression from the proposed mid-Holocene highstand at ca 3.9 cal. ka BP (Fig.  610   6). The elevation and timing of this proposed highstand fit well with previous evidence from 611 Ísafjarðardjúp, such as the raised beach surveyed by Principato (2008) at 5 m a.s.l. and dated to a.s.l.; 'modern'). Poor chronological control on the BB1 sample prevents a recent rate of RSL 621 change being calculated, although RSL must have risen to present from this saltmarsh peat 622 deposit, situated just below present sea-level.

Hrútafjörður-Heggstaðanes 624
The isolation basin records from Hrútafjörður-Heggstaðanes constrain the highest elevation 625 reached by postglacial RSL in the region. Despite the lack of direct evidence for a raised 626 shoreline, it is proposed that the marine limit lies between ~47 and 58 m a.s.l., with a minimum 627 timing for deglaciation of 11.2 cal. ka BP (Fig. 6). 628 The highest recorded marine influence along Hrútafjörður (~58 m a.s.l.; MY1) is higher than the 629 previously reported but undated marine limit in innermost Hrútafjörður at about 50 m a.s.l. 630 (Ingólfsson, 1991). However, previous research in northern Iceland has noted a southerly 631 decrease in marine limit elevations due to differences in deglacial timing (Norðdahl and Pétursson, 632 2005). In Hrútafjörður, it is likely that the marine limit formed during the Younger Dryas, based on 633 the known extent of the ice sheet during this period (e.g. Pétursson et al., 2015). As a result, the 634 marine limit features are assigned a tentative Younger Dryas age, given the lack of dateable 635 material to confirm the age of this feature. 636

Breiðavík-Látrar 637
The new late Holocene data from Breiðavík-Látrar provides a valuable constraint on recent RSL 638 changes in outermost Breiðafjörður, suggesting a rise in RSL since ca 1.4 cal ka BP. Investigation 639 of higher elevation basins in the region provides evidence for marine influence up to 67 m a.s.l.; 640 however, poor chronological control limits constraint of a regional RSL curve, likely as a 641 consequence of low productivity immediately following deglaciation, leading to low organic content 642 within dated bulk sediment samples. Investigation of the geomorphology of the area provides 643 support for a marine limit ~85 m a.s.l. (Norðdahl and Pétursson, 2005 The investigation of postglacial RSL change provides an opportunity to explore patterns of 648 postglacial uplift across NW Iceland. As outlined in Fig. 3 and Fig. 6, contrasting patterns of RSL 649 change would be expected in the four research locations studied under the two uplift scenarios. In 650 particular, the elevation of the marine limit is an important factor in establishing the most likely ice 651 loading/unloading scenario, which can be summarised as: