This package estimates channel parameters from Delft3D-based RESQML models. The repository is tightly linked with https://github.com/NorskRegnesentral/nrresqml

The main function is

channest.calculate_channel_parameters(settings, output_directory)

calculate_channel_parameters

Estimate channel parameters based on the provided parameters

settings

File path to a json file or a dictionary containing estimation settings. All settings are optional except data_file. In addition to these settings, advanced settings are described below. There are several available advanced settings. However, the default values have been determined experimentally and should work well for most Delft3D models. The advanced settings are documented below for completeness.

• data_file File path to a RESQML model (.epc file)

• crop_box Dictionary describing the extent of the model to use for estimation. Specified by providing keys x_0, x_1, y_0 and y_1 with float values. Delft3D models are typically starting at x=0, y=0.

output_directory

Directory to which output is written. The following files are written (relative to the provided directory):

• tw_scatter.png Scatter plot showing the channel thickness/width distribution per layer. Requires plotly-orca, otherwise, this is skipped.

• tw_scatter.html Scatter plot showing the channel thickness/width distribution per layer. Same as tw_scatter.png, except as html (based on plotly) which adds zoom and pan functions.

• summary.json JSON file containing the main results as well as the settings used to generate the results. Values under "channel-width" and "channel-height" are averaged over layers, with each layer having equal weight. Values under "segment-width" and "segment-height" are averaged over width segments, with each segment having equal weight.

Advanced settings

The advanced settings can be split in two: method-related and output-related. Some settings under method-related must be specified as lists of single values. All combinations of such values are then executed in a multi-configuration fashion, similar to vargrest. These settings are indicated by having a default values surrounded by [brackets].

Method-related parameters:

• merge_layers Number of layers to merge when calculating segments. Default is [5].

• alpha_hull Parameter to the alpha hull algorithm. 0.0 yields the convex hull. Default is [0.6]

• element_threshold Floating point threshold in number of layers for which points to include as channels in the merge layers. A value of None yields a default of including all points with a channel in at least one layer. Default is [None]

• mean_map_threshold Threshold between 0.0 and 1.0 used when filtering segments that cross areas not labeled as channel. A value of 1.0 removes all segments touching an area not labeled as channel. A value of 0.0 will only remove segments that does not touch areas labeled as channel at all. Default is [0.9]

• minimum_polygon_area Minimum area of the alpha polygon shape for it to be included in the estimation. Default is 100.

• turn_off_filters Disables all segment filters when set to True. Default is [False].

• step_z Sampling rate in z-direction in number of layers. Default is 1, which means all layers are sampled.

• z0 Starting layer for sampling in z-direction. Default is 0.

Output-related parameters:

• generate_plots Generate additional quality assessment plots. Default is False.

• generate_fences Generate poly lines as text files along the longest channel in each layer. These lines can be important and used as "fences" in RMS for. Default is False.

• pickle_data Store preliminary results in a Python pickle file. Main purpose is debugging or alternative post-processing. Default is False.

• scatter_max_width Length of the x-axis of the TW scatter plot, representing channel width. Default is 500.

• scatter_max_height Height of the y-axis of the TW scatter plot, representing channel thickness. Default is 14.