Cosmogenic nuclides

Figures from lsdtt3's cosmogenic tools: comparisons against CRONUS and published data, transient (step-change) and landslide experiments, and basin erosion-rate maps.

Chi and cosmogenic basins

Basins coloured by erosion rate (lsdviz3-cosmo-basins). Spatially joins point erosion rates from a cosmogenic CSV to basin polygons and draws the basins, coloured by erosion rate, over the standard hillshade, with the scaling, muon and atmosphere schemes in the caption:

pixi run lsdviz3-cosmo-basins --base runs/site_DEM.tif \
  --basins runs/site_junction_basin_polygons.fgb --cosmo-csv runs/site_erates.csv \
  --out figures/site_erates.png --colorbar

Flags: --base --basins --cosmo-csv --out [--hillshade --joined-basins-out --colorbar --scaling-scheme --muon-scheme --atmosphere --erate-input-units --allow-incomplete-basins]. Python: join_basins_with_cosmo_erates, render_cosmo_basin_erate_map.

Chi with erosion rates (lsdviz3-chi-cosmo). From a chi run's fit tables, writes the stacked and per-basin chi plots, a per-basin k_sn summary and map, and, given --cosmo-csv, k_sn against erosion rate with error bars:

pixi run lsdviz3-chi-cosmo --fit-nodes-csv runs/site_chi_fit_nodes_mn_0p450.csv \
  --fit-segments-csv runs/site_chi_fit_segments_mn_0p450.csv --outdir figures \
  --base runs/site_DEM.tif --basins runs/site_chi_basin_polygons.fgb \
  --cosmo-csv runs/site_erates.csv --theta 0.45

Flags: --fit-nodes-csv --fit-segments-csv --outdir [--base --hillshade --basins --basin-id-col --cosmo-csv --erate-col --lon-col --lat-col --sample-col --theta --scaling-scheme --atmosphere --muon-scheme --erate-input-units --allow-incomplete-basins --stem --dpi --cmap]. Python: render_chi_and_cosmo_basin_workflow, summarize_basin_ksn, render_chi_basin_ksn_map, charts.plot_ksn_vs_erate.

Cosmo comparison workflows

For cosmogenic workflows, lsdviztools3 can render standard comparison figures from lsdtt3 CSV artefacts.

Expected primary input:

• single_point_scan_comparison.csv, written by tests/scripts/run_cosmo_single_point_scaling_scan.py in a clone of lsdtt3 into the folder its --outdir names (it is generated, not tracked), for example:

cd ../lsdtt3
python3 tests/scripts/run_cosmo_single_point_scaling_scan.py \
    --pred-exe build/lsdtt-cosmo-prediction --tool-exe build/lsdtt-cosmo-tool \
    --with-cronus-v3 --sample-age-yr 0 \
    --outdir ../lsdviztools3/scan_out

--with-cronus-v3 queries the online CRONUS v3 calculator for the reference values that reference="cronus_v3" below compares against (without it those columns are empty), and --sample-age-yr 0 matches how CRONUS treats the samples.

The published-summary inputs below are tracked in lsdtt3, under tests/expected/cosmo/published/; the commands assume a clone of lsdtt3 next to this one.

Optional secondary input:

• sample-level comparison tables (for example OCTOPUS/riversand summary style CSVs containing a percent-difference column).

Python API:

from lsdviztools3.charts import load_cosmo_table, plot_cosmo_sweep_panels
from lsdviztools3.workflows import render_cosmo_comparison_figures

scan = load_cosmo_table("scan_out/single_point_scan_comparison.csv")
fig, axes = plot_cosmo_sweep_panels(scan, reference="cronus_v3", representative_pressure_hpa=900)

outputs = render_cosmo_comparison_figures(
    scan_comparison_csv="scan_out/single_point_scan_comparison.csv",
    sample_comparison_csv="../lsdtt3/tests/expected/cosmo/published/dethier2014_riversand_summary.csv",
    out_dir="demo_output/cosmo",
)
print(outputs)

Scripted example:

python examples/cosmo_comparison_workflow.py \
  --scan-csv scan_out/single_point_scan_comparison.csv \
  --outdir demo_output/cosmo

# Published-summary residuals (stone/lsdn/riversand)
pixi run lsdviz3-cosmo-published \
  --published-dir ../lsdtt3/tests/expected/cosmo/published \
  --datasets dethier2014,clapp2002,dibiase2010 \
  --mode auto \
  --outdir demo_output/cosmo

Main chart helpers:

• plot_cosmo_sweep_panels: nuclide x scaling panels of residual vs latitude, including separate muon-scheme traces.
• plot_cosmo_sample_residuals: grouped boxplots for sample-level residual tables, optionally split by scaling_scheme.

Main workflow helpers:

• render_cosmo_comparison_figures: sweep residual panels and optional sample-level residual figure.
• render_cosmo_published_residuals: dataset-level residual figure directly from <dataset>_summary.csv, <dataset>_lsdn_summary.csv, and <dataset>_riversand_summary.csv style files.

Transient cosmogenic and landslide charts

Charts over the lsdtt-cosmo-column-tool step_change_sweep CSVs (cosmo_transient_sweep_<label>.csv) and the per-sample <stem>_joined_samples.csv from a routed step-change landscape. Every quantity drawn is a column lsdtt3 already wrote: nothing is normalised or inverted here.

Function	What it draws
plot_cosmo_transient_concentration	Normalised concentration vs time since the step change, one series per nuclide (column prefix approach_frac_, norm_conc_ or norm_full_).
plot_cosmo_transient_erate_ratio	Apparent erosion-rate ratios vs time since the step change.
plot_nuclide_removal_response	Three panels explaining why stripping a surface moves the apparent ¹⁴C/¹⁰Be ratio.
plot_ratio_vs_steptime	Per-sample apparent-rate ratio vs time since the knickpoint passed, coloured by true basin-mean erosion rate.
plot_ratio_vs_mean_erate	The same ratio vs true basin-mean erosion rate, coloured by time since the step.
plot_ratio_overlay_1d	plot_ratio_vs_steptime with the equivalent 1-D column sweep drawn over it.
plot_scenario_delta	Paired change in the apparent-rate ratio, one series per scenario.
plot_landslide_area_volume_larsen	Realised landslide area–volume pairs (*_LS_INVENTORY.csv) against a Larsen-style power law.
plot_sample_effect_apparent_vs_true / plot_sample_effect_ratio	Landslide-vs-control apparent rates from *_LS_SAMPLE_EFFECT.csv.

Loaders: load_cosmo_transient_table, load_joined_samples, load_landslide_inventory. The sweep figures also have a CLI:

pixi run lsdviz3-cosmo-transient \
  --sweep-csv cosmo_transient_sweep_fast.csv cosmo_transient_sweep_slow.csv \
  --outdir demo_output/cosmo_transient \
  --families concentration erate_ratio