README.txt 

Supplementary Material for Paper 2014JB011078 
Circum-Arctic mantle structure and long-wavelength topography since the Jurassic.
Shephard, G. E.*1,2, Flament, N. 1, Williams, S. 1, Seton, M. 1, Gurnis, M. 3, and Muller, R.D. 1
Journal of Geophysical Research

*corresponding author: g.e.shephard@geo.uio.no
1.	EarthByte Group, School of Geosciences, University of Sydney, Sydney, Australia.
2.	Centre for Earth Evolution and Dynamics (CEED), Department of Geosciences, University of Oslo, Oslo Norway
3.	Seismological Laboratory, California Institute of Technology, Pasadena, CA 91125, USA.

############# Introduction ##############


This file and the associated Supplementary folder for download includes:
1. Supplementary Methods about the absolute reference frames and Net Lithospheric Rotation (NLR) involved in constructing models C1-C8. 
2. Supplementary References associated with these methods.
3. Two supplementary tables associated with the evolution of dynamic topography at the selected circum-Arctic locations (values used in-text) for C1 and C3-C5. Locations (coloured stars) identified on Figures 1, 6, 7 and 8 (S11 and S12 for alternative cases) and extracted from global grids of dynamic topography, as in Figure 6 and S11. One supplementary table with parameters for Cases C3-C8.
4. Fifteen Supplementary Figures and captions as referred to in-text
5. Link to Supplementary datasets including two alternative plate reconstructions (Shephard et al., 2013 and Seton et al., 2012) and forward model output including thermal structure and dynamic topography.

############# Supplementary tables ##############


Table S1 caption: Evolution of air-loaded dynamic topography (m) and its rate of change (m/Myr) at selected circum-Arctic locations through time for our preferred case C1. The time intervals between ~170, ~100, ~50 and 0 Ma were chosen to capture the main changes in dynamic topography trends and are to be used as a guide in conjunction with Figure 7. Note that shorter wavelength subsidence or uplift or changes in rates may occur within these intervals (see main text and Fig. 7). * denotes water-loaded values (Barents Sea and Lomonosov Ridge). Table can be used in conjunction with Figures 6-8 and values in-text. Blue text= subsidence, red text=uplift.

Table S2 caption: Evolution of air-loaded dynamic topography (m) and its rate of change (m/Myr) at selected circum-Arctic locations through time for three alternative cases, separated by commas in order of C3, C4, C5. The time intervals between ~170, ~100, ~50 and 0 Ma were chosen to capture the main changes in dynamic topography trends. Note that shorter wavelength subsidence or uplift or changes in rates may occur within these intervals (see main text and Figs. 7 and S11, S12). * denotes water-loaded values (Barents Sea and Lomonosov Ridge).  Table can be used in conjunction with Figures S11-S12 and values in-text. Blue text= subsidence, red text=uplift.

Table S3: Acronyms and alternative model details referred to in this study.

############# Supplementary Figure captions ##############
Figure S1. 180-30 Ma evolution of the plate reconstruction (Shephard et al., 2013 with a modified reference frame, Table 2) assimilated in the mantle flow models. The absolute reference frame used here is that of case C1. Reconstructed plate boundaries (black lines with teeth located on overriding plate), coastlines (dark grey lines), continental lithosphere (grey polygons) and ages of oceanic lithosphere (see colour scale) are shown, as well as velocities (black arrows). Major plates and oceans labeled as AM Amerasia Basin, AFR Africa, CCR Cache Creek oceanic plate, EUR Eurasia, GRN Greenland, FAR Farallon, IZA Izanagi, MOK Mongol-Okhostk, NAM North America, SAO South Anuyi Oceans. Orthographic projection centered on 30¡W. Additional reconstruction ages are shown in Fig. 2.

Figure S2. Top panels, maps of predicted time-dependent temperature field from case C1 at 1000 km depth. Bottom panels, maps of predicted present-day temperature field for case C1 at different depths from 500 km to the near the core-mantle boundary (CMB ~2900 km). Slabs labeled as in text and Figs. 3-5. Present-day coastlines superimposed in black for reference. Cold material (T < 0.45) is inferred to represent subducted lithosphere whereas hot material represents upwelling from the thermal boundary layer along the CMB.

Figure S3. Predicted present-day mantle temperature field for cases C3-C5 at different depths from 500 km to near the core-mantle boundary. Present-day coastlines superimposed in black for reference. Cold material (T < 0.45) is inferred to represent subducted lithosphere whereas hot material represents upwelling from the thermal boundary layer along the CMB. Slabs labeled as in text and Figs. 3-5 and S6-S8, S10.

Figure S4. Predicted time-dependent mantle temperature for cases C3-C5 at 1000 km depth. Present-day coastlines superimposed in black for reference. Cold material (T < 0.45) is inferred to represent subducted lithosphere whereas hot material represents upwelling from the thermal boundary layer along the CMB.

Figure S5. Predicted time-dependent mantle temperature for cases C1 and C2 (Table 2) and comparison to seismic tomography for the present-day. As in Figure 3 but for a cross-section at 50¡N latitude across NAM (130-30¡W). Inferred slabs from this vertical cross-section result from subduction along the north-eastern margin of Panthalassa (a, c, d) and along the intra-oceanic subduction zone of the Wrangellia Superterrane (b). 

Figure S6. Predicted evolution of mantle temperature for cases C3, C4 and C5 (Table S3) and comparison to seismic tomography for the present-day. Top panels, orthographic projection of cross-section at 50¡N latitude across NAM (130-30¡W) superimposed on location of subduction zones and predicted present-day temperature at ~1500 km depth. Inferred slabs from this vertical cross-section correspond largely to subduction along the north-eastern margin of Panthalassa and along the intra-oceanic subduction zone of the Wrangellia Superterrane. Panels in green box show seismic velocity anomalies for three tomography models with 0.45 mantle temperature contours overlain for cases C3 (green), C4 (black) and C5 (purple).


Figure S7. Predicted time-dependent mantle temperature for cases C3, C4 and C5 (Table S3) and comparison to seismic tomography for the present-day. As in Figure S6 but for a cross-section at 30¡N latitude across NAM.

Figure S8. Predicted time-dependent mantle temperature for cases C3, C4 and C5 (Table S3) and comparison to seismic tomography for the present-day. As in Figure S6 but for a cross-section at 40¡W latitude under Greenland (40-90¡N). At 150 Ma, two subducting slabs are captured in case C3, 75¡N and 85¡N, and a single slab at 75¡N is clearly imaged in cases C4 and C5 with a second smeared slab under 85¡N. The difference in location and dip of subducting slabs at this fixed vertical cross-section is a function of absolute reference frames used (Table S3). 

Figure S9. Predicted evolution of mantle temperature for cases C1 and C2 (Table S3) and comparison to seismic tomography for the present-day. As in Figure 3 but for a cross-section at 60¡N longitude under Siberia (90-180¡E). Inferred slabs within this cross-section correspond largely to subduction of the Izanagi Plate along the north-western margin of Panthalassa (p). 

Figure S10. Predicted time-dependent mantle temperature since initial conditions for cases C3, C4 and C5 (Table S3) and comparison to seismic tomography for the present-day. As in Figure S6 but for a cross-section at 60¡N latitude under northern Eurasia (0-100¡E). Inferred slabs within this cross-section largely result from subduction along the northern margin of the Mongol-Okhotsk Ocean (m) and along the northwestern margin of Panthalassa (p). Note that case C3 has an initial slab depth to 1750 km (Table S3) as opposed to the alternative cases, which are to 1210 km.

Figure S11. Air-loaded surface dynamic topography for cases C3-C5 between 170-0 Ma, as in Figure 6. Stars indicate location of selected reconstructed Arctic points as in Fig. 1. Orthographic projection centered on 30¡W. 

Figure S12. Predicted evolution of dynamic topography for cases C3-C5 at selected circum-Arctic locations grouped into four geographic regions between 170-0 Ma (based on the plate reconstruction). The colours of the plotted lines match the colours of the stars in Fig.1, and solid for C3, thick for C4 and dashed for C5. Note the broad subsidence predicted for most locations from 170 Ma to between ~70-50 Ma followed by slowed subsidence or uplift to present day. Values are detailed further in Table S2. Air-loaded results shown for all locations except for Lomonosov Ridge and Barents Sea which are water-loaded.

Figure S13. Evolution of Net Lithospheric Rotation (NLR) for the five main reconstructions used herein (Table 2, S3), and present-day NLR calculated from reference frame HS3, based on Pacific hotspots (0.44¡/Myr). NLR evolution was computed in 1 Myr increments, which is the interval at which boundary conditions are defined for the geodynamic models. Conrad and Behn (2010) proposed that 60% of HS3 (0.26¡/Myr) is the geodynamically reasonable limit for NLR. Larger NLR from the reconstructions likely reflects the motion of large, fast-moving plates of Panthalassa, for which the reconstruction uncertainty is large before 83.5 Ma. NLR computed using the same relative plate motions as in Seton et al. (2012) and the absolute reference frame of Doubrovine et al. (2012) is shown for reference in green. The peak amplitudes at ~80 Ma for DBV is larger than in Fig. 9 of Doubrovine et al. (2012) that showed NLR computed in 10 Myr incrmenets. Other small differences may also arise due to different Pacific plate boundaries and the use of a Pacific plate circuit via Antarctica (Seton et al., 2012; Shephard et a;., 2013) rather than via the Lord Howe Rise (Doubrovine et al., 2012).

Figure S14. Left panels, predicted present-day mantle temperature for cases C1-C8 (Tables 2, S3) and comparison to seismic tomography, middle panels. Details are as in Figs. 3 and S5, this location is at 50¡N latitude across NAM (130-30¡W). Right panels show the evolution of dynamic topography for the North American region illustrating the similarity between models despite variation in the predicted lower mantle structure.

Figure S15. As in Figure S14. Left panels, predicted present-day mantle temperature for cases C1-C8 (Tables 2, S3) and comparison to seismic tomography for the present-day, middle panels. This location is at 60¡N latitude across Eurasia (0-100¡E). Right panels show the evolution of dynamic topography for the Barents Sea region illustrating the similarity between models despite variation in the predicted lower mantle structure.



############# Supplementary Datasets ##############

The associated data for the plate reconstruction of Shephard et al. (2013) can be downloaded at 
ftp://ftp.earthbyte.org/papers/Shephard_etal_Arctic_plate_model 
and for the Seton et al. (2012) model at 
ftp://ftp.earthbyte.org/papers/Seton_etal_Global_ESR/Seton_etal_Data.zip

Raster images (jpg stacks, age/depth-coded for loading into GPlates) of model case "C1" including
1. Air-loaded surface dynamic topography at selected Myr increments (e.g. 0Ma, 20Ma etc.), as in Figure 6. Colour palette provided. and
2. Depth dependent mantle temperature structure (at present-day but depth-age coded), as in Figures 3-5, S2. Colour palette (non-dimensional temperature) provided.
These files can be downloaded at ftp://ftp.earthbyte.org/papers/Shephard_etal_Arctic_mantle/Shephard_etal_2014.zip

These files (.gpml and shapefile extensions) can be loaded into the freely-downloadable desktop plate tectonic reconstruction software GPlates: www.gplates.org
See the respective README for more file-specific information

G.Shephard
12.06.2014