Landscape evolution

Animations of lsdtt-lem runs, and the figures used to test the hillslope solvers against analytical and benchmark solutions.

Animating a transient run (lsdviz3-animate)

pixi run lsdviz3-animate --in runs/lem --out figures/lem.gif --fps 8

Turns the _model_iteration_<N> rasters of an lsdtt-lem run (written every frequency_of_output_writing steps) into a GIF, annotated with model time from _model_elevation_stats.csv. Flags: --in --out --glob --stats-csv --fps --loop --no-annotate --keep-frames, plus the shared style flags.

Incision pulses (lsdviz3-lem-pulse)

pixi run lsdviz3-lem-pulse --in runs/pulse --prefix pulse --out figures/pulse.gif

Each frame is a hillshade with the erosion rate draped over it, the channel network, and a panel of the run's history. Needs a run with write_erosion_rate_raster and write_channel_mask_raster. Flags: --in --out --prefix --ext --stats-csv --erosion-cmap --erosion-vmin --erosion-vmax --erosion-percentile --erosion-transparency --channel-color --no-timeseries --no-annotate --fps --loop --keep-frames, plus the shared style flags.

Channelization-index ladders (lsdviz3-lem-ci)

pixi run lsdviz3-lem-ci --in runs/ci --out figures/ci.gif --view oblique

Animates a channelization-index ladder run: the CI history panel beside a map or an oblique view of the landscape. Flags: --in --out --view --prefix --ext --stats-csv --no-initial --map-width --azimuth --elevation --vertical-exaggeration --landscape-cmap --fps --frame-dpi --loop --keep-frames, plus the shared style flags.

Python: animate_transient_rasters, animate_lem_pulse and animate_ci_run, with the frame finders find_transient_frames, find_pulse_frames and find_ci_frames.

Hillslope solver comparisons

Figures comparing lsdtt3's hillslope solvers (Newton and cellular) with each other and with analytical solutions. Each module docstring explains what the figure tests and how to read it; read it before quoting a result.

Function	Module	What it draws
plot_estar_rstar_model_scan	charts.estar_rstar	Analytical, Newton and cellular strip results in E–R space.
plot_hillslope_profiles / plot_profile_residual_panels	charts.hillslope_profiles	Elevation, gradient, curvature and residual panels for 1-D hillslope comparisons.
plot_residual_density / plot_residual_raster	charts.hillslope_residuals	Residual histograms (optionally hillslope-masked) and residual maps centred on zero; summarise_residuals tabulates RMS/L∞/percentiles.
plot_transient_residuals / plot_strip_transient_metrics	charts.hillslope_transient_residuals	RMS and max residuals through time; relief, max gradient and divide Laplacian changes per model.

# paired Newton vs cellular hillslope solver animation
pixi run lsdviz3-hillslope-comparison --newton runs/newton --cellular runs/cellular \
  --out demo_output/hillslope_comparison.gif --layout 2d

The animation is built by render.hillslope_comparison_animation.render_hillslope_comparison_animation; examples/hillslope_comparison.py drives the whole comparison workflow.

The Perron and Ren benchmark figures are made by examples/perron_ren_p1_p2.py … perron_ren_p6.py (P1–P2 outputs, P3 timestep independence, P4 transient convergence and runtimes, P5 stability envelope, P6 spatial convergence and anisotropy).