Contents

Reference

Reference

glaft.Velocity class

glaft.Velocity(vxfile: str=None, vyfile: str=None, wfile: str=None, static_area: str=None, on_ice_area: str=None, nodata: float=-9999.0, velocity_unit: str='m/day', thres_sigma: float=2.0, kde_gridsize: int=60)

Construct an experiment to calculate velocity map benchmarking metrics.

vxfile:        str, Vx raster file path 
                    (single band with meters as the geotransform unit)
vyfile:        str, Vy raster file path 
                    (single band with meters as the geotransform unit)
wfile:         str, weight raster file path [Optional]
static_area:   str, static area (shapefile) file path
on_ice_area:   str, on-ice area (shapefile) file path
nodata:      float, nodata value in the provided geotiff. NOT FULLY IMPLEMENTED YET.
velocity_unit: str, velocity unit to be shown on the result plots.
thres_sigma: float, selected thresholding z values.
kde_gridsize:  int, grid size used for evaluating a crude KDE surface 
                    (larger value means a faster but less precise process)

Velocity.static_terrain_analysis(plot=None, ax=None)

Perform the static terrain analysis and calculate the correct-match uncertainty.

plot: None for no plot; 
      "full" for plotting the results in the full extent; 
      "zoomed" for plotting the results in the zoomed extent.
ax:   matplotlib.axes object to place the plot.

Velocity.longitudinal_shear_analysis(plot=None, ax=None)

Perform the longitudinal strain rate analysis and calculate the associated metrics.

plot: None for no plot; 
      "full" for plotting the results in the full extent; 
      "zoomed" for plotting the results in the zoomed extent.
ax:   matplotlib.axes object to place the plot.

Velocity.plot_full_extent(ax=None, rect=None, metric: int=1, **pt_style)

Plot the analysis results in the full extent.

ax:       matplotlib.axes object to place the plot.
rect:     style of the rectanglar box indicating the correct match area. 
          If None, the plot uses thick red line as default. 
          See Velocity.create_rectangle_patch for details.
metric:   which metric to be plotted. 
          1 for static terrain; 
          2 for along-flow strain rate.
pt_style: point style passed to plt.scatter.

Velocity.plot_zoomed_extent(ax=None, rect=None, base_colormap=None, metric: int=1, **pt_style)

Plot the analysis results in the zoomed extent focusing on the correct-match area.

ax:       matplotlib.axes object to place the plot.
rect:     style of the rectanglar box indicating the correct match area. 
          If None, the plot uses thick red line as default. 
          See Velocity.create_rectangle_patch for details.
base_colormap: colormap for showing the KDE probability distribtuion. 
          If None, the default (cramericm.bamako_r) is used. 
metric:   which metric to be plotted. 
          1 for static terrain; 
          2 for along-flow strain rate.
pt_style: point style passed to plt.scatter.

Velocity.create_rectangle_patch(var_x, var_y, **rect_style)

Create a rectangle patch to be plotted on the visualization of the processing results.

var_x: half width of the rectangle
var_y: half height of the rectangle
rect_style: style of the rectangle. 
            Default is {'linewidth': 2, 'edgecolor': 'xkcd:cranberry', 
            'facecolor': 'none', 'alpha': 0.7}

Returns
--------------
matplotlib.patches.Rectangle object.

Velocity.plot_strain_map(ax=None, base_colormap=None, vmax=None)

Show the strain map. Only works after longitudinal_shear_analysis is performed).

ax:       matplotlib.axes object to place the plot.
base_colormap: colormap for showing the strain map. 
          If None, the default (cramericm.bamako) is used. 
vmax:     value at where color is saturated. 
          If None, the saturation limit will be automatically determined.

Velocity.cal_invalid_pixel_percent()

Calculate the amount of invalid pixels in a velocity map.

Important attributes

  • Velocity.vxfile Vx file path.

  • Velocity.vyfile Vy file path.

  • Velocity.wfile Weight file path.

  • Velocity.static_area static area geometries file path.

  • Velocity.on_ice_area on-ice area geometries file path.

  • Velocity.nodata NoData value for the raster files.

  • Velocity.velocity_unit velocity unit to be shown on the result plots.

  • Velocity.vx 2-D, clipped vx data

  • Velocity.vy 2-D, clipped vy data

  • Velocity.xy = np.vstack([flattened_vx, flattened_vy]) –> (1-D)-by-2 array containing vx and vy.

  • Velocity.w 2-D, clipped weight data

  • Velocity.w_flat 1-D, clipped & flattened weight data

  • Velocity.dx velocity map pixel spacing, x

  • Velocity.dy velocity map pixel spacing, y

  • Velocity.thres_sigma selected thresholding z values.

  • Velocity.kernel KDE kernel (default epanechnikov)

  • Velocity.bandwidth KDE bandwidth (default: calculated using the rule of thumb)

  • Velocity.kde_gridsize grid size used for evaluating a crude KDE surface (larger value means a faster but less precise process)

  • Velocity.mesh_fine Final KDE mesh

  • Velocity.mesh_fine_z KDE values at the final KDE mesh vertices

  • Velocity.mesh_fine_thres_idx boolean array showing whether a vertex falls within the correct match area

  • Velocity.metric_static_terrain_x delta_x

  • Velocity.metric_static_terrain_y delta_y

  • Velocity.kdepeak_x KDE peak location x

  • Velocity.kdepeak_y KDE peak location y

  • Velocity.outlier_percent Incorrect match percentage (*100%)

  • Velocity.invalid_percent Invalid pixels percentage (*100%)

  • Velocity.flow_theta Flow direction

  • Velocity.exx normal strain rate exx, image axis

  • Velocity.eyy normal strain rate eyy, image axis

  • Velocity.exy shear strain rate exy, image axis

  • Velocity.flow_exx normal strain rate exx, flow axis (that is, ex’x’)

  • Velocity.flow_eyy normal strain rate eyy, flow axis (that is, ey’y’)

  • Velocity.flow_exy shear strain rate exy, flow axis (that is, ex’y’)

  • Velocity.metric_alongflow_normal delta_x’x’

  • Velocity.metric_alongflow_shear delta_x’y’

Auxillary functions

glaft.show_velocomp(gtiff: str, ax=None, **cm_settings)

Preview a geotiff file (presuming a velocity map).

Parameters
-------------
gtiff: geotiff file path (much be single band. e.g., Velocity component, such as Vx)
ax:    matplotlib.axes object to place the plot
cm_settings: colormap settings to be used in the visualization. 
       Default: {'vmin': -2, 'vmax': 2, 'cmap': cramericm.broc_r}

Returns
--------------
cm_settings: colormap settings using in the visualization.

glaft.prep_colorbar_mappable(vmin=-2, vmax=2, cmap=cramericm.broc_r)

Generate a mappbale object for creating a colorbar.

Parameters
-------------
vmin: colormap lower range
vmax: colormap upper range
cmap: colormap object as the master colormap

Returns
--------------
mappable: matplotlib.mappable object

glaft.naof2(im)

NAOF preprocessing filter. Translated to Python from the GIV source code. All implementation credit to Max VWDV.

MaxVWDV. (2021). MaxVWDV/glacier-image-velocimetry: Glacie Image Velocimetry (v1.0.1). Zenodo. https://doi.org/10.5281/zenodo.4904544

Parameters
-------------
im:       np.ma.array;  input image

Returns
--------------
naof_im:  np.ma.array;  image with NAOF filter applied

previous

GLAFT Quick Start

next

Table S1: Parameters of all 172 tests