Through our calculation for GlacierMIP3 experiment, we have noticed some mismatch among elevation information in RGI6. By sharing this issue with you, we hope that you could create more accurate data in RGI7.
What we did:
We compared elevations of terminus, median, and top of individual glaciers in the attribute data with those extracted from the hypsometry, both provided as RGI6. We made scatter plots and linear regression lines for the 19 RGI regions (an attached figure). Despite the hypsometry based elevations at a 50-m elevation interval, the overall relationships seem very good (it should be very good of course). However, the plot seems scattered for the region 19 Antarctica. and some terminus elevations mismatch. In addition, the elevations in the region 16 Low Latitudes are systematically biased (Attribution z > Hypsometry z). We think that this is due to different DEMs used for each data.
Figure Caption: Comparison of top (red), median (yellow), and terminus (blue) elevations of individual glaciers in the 19 RGI6 regions. Horizontal and vertical axes denote the elevations in attribute data and hypsometry data, respectively. Colored texts denote the linear regressions of the three elevations (y = ax + b, unit is m).
RGI version RGI v6
Describe the issue
Email by Koji and Akiko received Jan 13, 2022:
Through our calculation for GlacierMIP3 experiment, we have noticed some mismatch among elevation information in RGI6. By sharing this issue with you, we hope that you could create more accurate data in RGI7.
What we did: We compared elevations of terminus, median, and top of individual glaciers in the attribute data with those extracted from the hypsometry, both provided as RGI6. We made scatter plots and linear regression lines for the 19 RGI regions (an attached figure). Despite the hypsometry based elevations at a 50-m elevation interval, the overall relationships seem very good (it should be very good of course). However, the plot seems scattered for the region 19 Antarctica. and some terminus elevations mismatch. In addition, the elevations in the region 16 Low Latitudes are systematically biased (Attribution z > Hypsometry z). We think that this is due to different DEMs used for each data.