Multi-temporal analysis

Once a set of scenes of the same reach has been processed (see batch processing), rivabar can split the rivers at persistent tributary confluences and measure migration between scene pairs.

Splitting at confluences and matching segments

import rivabar as rb

# 1. Cluster tributary confluences detected across scenes
common_confluences = rb.find_common_confluences(rivers, min_scene_fraction=0.5)

# 2. Split each river's main path at the common confluences and group
#    corresponding segments across scenes
segment_groups, rejected = rb.match_river_segments(rivers, common_confluences)

# 3. Visualize the matched segments
rb.plot_river_segments(rivers, common_confluences, segment_groups=segment_groups)

Curvature–migration analysis

Pairwise DTW-based analysis of migration rates, either on full main paths or per segment group:

# full reaches
results = rb.analyze_river_pairs_filtered(rivers, min_width_m=100,
                                          min_time_gap_days=365,
                                          max_time_gap_years=5)

# or per segment group, merged into one DataFrame
results_list, df = rb.analyze_all_segment_groups(segment_groups)

The output DataFrame contains, per scene pair, the curvature–migration rate correlation (r_curv_mr), the regression slope in physical units (slope_curv_mr), width-regression diagnostics (width_slope, a stage indicator), and a pair_class label (good / stage_mismatch / reversed / weak / insufficient) from classify_pairs.

Use rb.interactive_scatter(df) to select pairs in a notebook and rb.plot_pair to inspect a single pair with filled channel polygons and centerlines.

Deposition and erosion maps

fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(20, 15), sharex=True, sharey=True)
chs, bars, erosions, dates, centerlines = rb.create_and_plot_bars(
    rivers, 0, len(rivers) - 1, ax1=ax1, ax2=ax2)