Glacier morphometry

A set of R scripts to facilitate geomorphometric analyses of glaciers.

Installation

Requirements

The scripts were written and tested in the following environment: R version 3.4.4 (2018-03-15) Platform: x86_64-redhat-linux-gnu (64-bit) Running under: CentOS Linux 7 (Core)

Required R packages are: rgdal, raster, ggplot2, tidyr, dplyr, gridExtra and e1071, as well as their dependencies.

Getting started

Part 1: Preparations

Download and unpack ALOS tiles Download and unpack RGI. Combine all relevant glacier SHPs into a single SHP for input. List the glaciers of interest in input.csv.

Part 2: Create glacier DEMs

Change definitions as wished in specifications.R.

Run Rscript main.R.

Part 3: Analyze glacier DEMs

Change settings in analysis_parameters.R as needed.

Run Rscript analyze.R.

Outputs

Plateau detection

Classified result raster

value   meaning
_____   __________________________________________________
    0   glacier
    1   flat spots on glacier
  100   plateau elevation band
  101   flat spot on plateau elevation band, but too small
  111   plateau (flat spot, big enough)
   NA   background

Plateau elevation raster

Metrics

Two metrics, the min and max elevation of plateaus are directly derived within the detection function and are added to the output. The definition and all externally derived outputs can be foud in the Metrics definition section.

Metrics definitions

General metrics

area_absolute_glacier

The area covered by the glacier raster, in square meters. Calculated using the raster; by number and dimension of cells dimensions.

elevation_min, elevation_max, elevation_mean

The minimal, maximal and arithmetic mean elevation of the glacier raster.

elevation_range

The range from the minimal to the maximal elevation of the glacier raster; maximal elevation minus minimal elevation.

ela_calculated

The Equilibrium Line Altitude (ELA), calculated using the glacier raster and the assumed relative elevation of the ELA, as set as Input in the analysis_parameters.R file.

skewness

The skewness of the glacier raster's elevation distribution, calculated using b_1 = m_3 / s^3 = g_1 ((n-1)/n)^(3/2) by e1071::skewness. All three approaches that the function provides have been tested and showed only minimally different results, mainly in the decimal places of the results.

Plateau detection derived metrics

pleateau_elevation_min, plateau_elevation_max, plateau_elevation_range

The lowest and highest elevation, and the range in between, of the detected plateaus. Derived within the plateau detection function itself, saved in the processing loop.

plateau_elevation_mean, plateau_elevation_sd, plateau_elevation_skewness

The arithmetic mean, the standard deviation and the skewness of the plateau elevations. Derived in the processing loop, after the plateau detection execution, from the result. The skewness is calculated in the same way as for the entire glacier, as described in previous sections.

area_absolute_*

The absolute area in square meters. If a cell fits into different categories, it is counted in all of them.

area_absolute_glacier is not derived in the plateau detection itself, but before in the main metrics script. It is described in the previous section.

area_absolute_flat is the sum of all flat areas on the glacier (with no minimum area), that do not exceed the threshold

Acknoledgements

These scripts were developed within the research project "TopoClimatic Forcing and non-linear dynamics in the climate change adaption of glaciers in High Asia" (TopoClif). TopoCliF and the work of David Loibl and Arne Thiemann within the project were funded by DFG under the ID LO 2285/1-1.