Versions of the inventory: v1: made at Scripps, final version 16/10-2013 v2: some updates at NP, final version 24/4-2014 v3: cleaned up and added some more, final version 27/3-2015 v4: some smaller updates and full ststistics calculation, final version 24/4-2015 v5: added more visible rumples and a few promotories in East Ant, final version 1/9-2015 ------------ updates v5 - a few corrections from Kenny - checked consistency with ADD land classifications, a few adjustments - added 2 undetected rumples from Forste et al. (2015) - added about 50 more rumples from visual inspection in MOA and LIMA (hills, flow scars) - new ice rise id numbers based on sorted centroid longitudes (id 1-704) ------------ updates v4 - a few corrections from Kenny - checked against the SCAR name database in Quantarctica - ran full statistics calculation to extract final attributes - this is the final version for the submitted paper ------------ updates v3 - made corrections from Kenny's list - added Cesar's ice rumples in DML - adding palce namse from online database - Checked promotories around the continent and added the most relevant ones - no updated statistics ------------ Updates v2 - added more promotories - added back ice rises/islands that are grounded over sea level (483 -> 598 features) (but some added from MOA products -> not always consistent with Scripps grounding lines!) - added several attributes and sub-files with relevant data: icerises_inventory_v2 - base file with similar attributes as v1 (598 features) icerises_inventory_v2_attr - same as over, but with many additional attributes generated in matlab icerises_inventory_v3_combined_pts - 900 m grid points with extracted attributes from other grids (used by matlab to generate extra attributes in the _attr shapefile) I think I have now extracted most of the parameters we talked about. The way I did it was that I gridded the inventory at 900 m resolution and then converted it to a point data set covering all icerise polygons (named ‘icerises_inventory_v2_combined_pts’) . For this point data set, I extracted velocity (Rignot et al), bed and surface elevation (from bedmap 2), and surface slope (calculated from bedmap2). Then I looped through all ice rise polygons and calculated parameters such as mean velocity, min and max elevation, mean slope etc. These statistics were then added to the v2 inventory as extra attributes and a new shapefile (‘icerises_inventory_v2_attr’). In case you want to do something on your own (or Cesar?) or adjust any of the plots, I put all shapefiles with corresponding mat-files and scripts here: https://dl.dropboxusercontent.com/u/32616939/for_kenny.zip I tried to plot various histograms from the additional attributes, but only area and velocity gave me something I could easily interpret (attached as PNG images). It is relatively easy to plot other parameters with the matlab script if you want to explore something, load the mat-file instead of shape/dbf if you get format problems. I think the attributes are pretty intuitive, if not you can quickly check the calculations in the other matlab script. ‘Width’ and ‘length’ are the shortest and longest distances (in unit km) from the summit to the edge of each ice rise/rumple. I hope this is okay for now, and then we can clean up and quality control everything before publication. One additional analyses we talked about that I haven’t gotten around to do is looking at potential relations between ice shelf velocity and proximity to ice rises (i.e. ice rumple velocity vs. farfield velocity). This should be possible, but requires some more work. I am not sure how interesting those results will be though. It could potentially be more useful to look at more idealized cases along ice shelf flowlines, e.g. the Doake ice rumples on Filchner-Ronne. I have made some files with FRIS streamlines where the velocity development across rumples can be easily plotted: https://dl.dropboxusercontent.com/u/32616939/FRIS_stream_5km_5yr.zip ------------ Icerises inventory v1 This inventory of ice rises and rumples is compiled from a combination of the MODIS Mosaic of Antarctica (MOA), the Landsat Image Mosaic of Antarctica (LIMA), and groundingline points from ICESat altimetry and SAR Interferometry [Scambos et al., 2007; Haran et al., 2005; Bindschadler et al., 2011; Fricker et al., 2009; Brunt et al., 2010; Rignot et al., 2011;]. The map excludes features where the majority of the bedrock elevation is estimated to be above sea level in the Bedmap2 data set [Fretwell et al., 2012] although they may look similar visually. Polygon attributes: type: 1 = Ice rise (topographic feature within an ice shelf, has its own dynamics) 2 = Ice ridge (major ridges between ice shelves/streams with a clear divide) 3 = Ice rumple (smaller irregular grounded areas with limited prominence and typically ice-shelf flow across it) divide: 0 = no clear divide 1 = clear divide island: 0 = grounded below sea level 1 = partly grounded above sea level isolated: 0 = inside or connected with an ice shelf 1 = not connected to an ice shelf (only applies to Siple Coast) name: only for the major ice rises so far MOA2003, MOA2009, InSAR (main data sources): 0 = not represented in that data set 1 = represented in that data set (hence three zeros means manual identification)