Comparision of calculated and measured paleo-sea level proxies with PaleoMIST 1.0 version 2.0 As a supplement to “A new global ice sheet reconstruction for the past 80 000 years” by Evan J. Gowan, Xu Zhang, Sara Khosravi, Alessio Rovere, Paolo Stocchi, Anna L. C. Hughes, Richard Gyllencreutz, Jan Mangerud, John-Inge Svendsen & Gerrit Lohmann Evan James Gowan These reports were generated using the Global Archive of Paleo Sea Level Indicators and Proxies (GAPSLIP). The sea level data, plus the scripts used generate the report, can be found on Github: https://github.com/evangowan/paleo_sea_level In this package, there are three reports, each of which have six model runs that are compared with the standard PaleoMIST 1.0 ice sheet margins, ice sheet and paleo-topography reconstruction. In contrast to previous versions of this report, the standard version is the MIS 3 scenario where Hudson Bay remains ice covered. All models have been computed at 500 year resolution by linearly interpolating the ice load between time steps. This was done because the original 2500 year resolution overestimates the loading effect in near field regions due to the fact that SELEN uses a heaviside function to represent the load. The first document (report1) shows a comparison of the response using a variety of lower mantle viscosity values. The second document (report2) compares the standard PaleoMIST 1.0, the alternate reconstruction with an ice free Hudson Bay in MIS 3 (in prior versions of this report, this was the default) a version with Antarctica set to be in its present configuration at 5000 years ago, ICE-6G, and PaleoMIST with three other commonly used Earth models. The third document (report3) includes several extra models where the thickness of the Greenland Ice Sheet was increased, and the standard PaleoMIST with different lithosphere thickness values. I have also included the explicitly calculated sea level for all the models displayed in these reports in a zip file called "calculated_sea_level.zip". You can place the files in the folder "latex/calculated_sea_level" after you download the Git repository. A summary of the three documents are below: ------------------------------------- report1 - lower mantle models.pdf ------------------------------------- In this report there is a detailed summary, including plots, of a worldwide compilation of paleo-sea level data, and six ice sheet-Earth models. In this particular report, we compare the standard version of PaleoMIST 1.0 (with 2500 year time steps and using a lower mantle viscosity of 4 × 10^22 Pa s), with five other Earth models with viscosity values ranging between 10^21 and 10^23 . When developing PaleoMIST 1.0, a variety of lower mantle viscosity values were tested, and it was found that a value approaching 10^23 Pa s provided the best trade-off between increasing the amount of ice in the center of the Laurentide Ice Sheet and fitting the sea level data. This ended up being true for the Eurasian ice sheets as well. PaleoMIST 1.0 was tuned to an Earth model with a viscosity of 4 × 10 22 Pa s, but the comparison shown in this document demonstrate that a slightly higher value of 10^23 Pa s provides an even better fit. ------------------------------------- report2 - several models.pdf ------------------------------------- Note, the models in this report are different than in previous versions. In this report there is a detailed summary, including plots, of a worldwide compilation of paleo-sea level data, and six ice sheet-Earth models. In this particular report, we compare the standard version of PaleoMIST 1.0 (with 2500 year time steps and using a lower mantle viscosity of 4 × 10^22 Pa s), an alternative version of PaleoMIST 1.0, with an ice free Hudson Bay during MIS 3, and an alternative model where the Antarctica ice sheets have modern ice thickness from 5000 years ago. The standard version of PaleoMIST 1.0 is calculated using Earth models proposed by James et al. (2009) (i.e. for tectonically active areas), VM5a from Peltier et al. (2015), the optimal model from Lambeck et al. (2017). Finally, I have also included ICE-6G VM5a for comparison. Which of the two different scenarios for MIS 3 is more likely cannot be discriminated with the available data. Comparing the standard version of PaleoMIST 1.0 with other Earth models utilized in other studies show that our chosen Earth model provides a better fit to the data in formerly glaciated areas. This is unsurprising, since the ice model was tuned to our chosen Earth model. The alternative Antarctica model improves the fit of far field sea level during the Holocene. ------------------------------------- report3 - Greenland ice sheet and lithosphere thickness changes.pdf ------------------------------------- This report tests shows tests of a few Greenland models with additional ice thickness in an attempt to improve the fit to the data added in GAPSLIP 2.0. The results show that adding extra ice does not substantially improve the fit, so it is likely that the margin chronology needs to be adjusted in PaleoMIST. I have also included models with 60 km and 90 km lithosphere (default is 120 km), which does improve the fit in some areas in Greenland. --------- Version 1.1: In this version, I have recalibrated all of the data with the new IntCal20 curves. I have added a few new LGM sites. I have also restored some of the data that I conservatively set to be marine limiting due to large uncertainties back to being index points. --------- Version 1.2: This version includes an update of the Baltic Sea dataset, as well as data from the North Sea region in Europe. --------- Version 1.3: Added Antarctica data for the Holocene and MIS 3, improved the presentation of index points on the plots, added the ability to calibrate mixed dates and terrestrial reservoir corrections, and improved documentation. --------- Version 2.0: I have included data from Greenland and Australia. I have changed the default version of PaleoMIST to be the MIS 3 ice covered Hudson Bay scenario, which I think is more likely. I have changed some of the models in the second report (as listed above) and included a third report with alternative Greenland models.