GCS nesting analysis outputs
Methodology + motivation
Cross-sections along each of the selected flow stage centerlines are aligned to the longest centerline path (typically the lowest flow stage) based on Euclidian distance.
Ws, Zs, C(Ws, Zs) values are aligned for all flow stages, faciliating comparison.
GCS nesting analysis attempts to capture the nested hierarchical topographic structure of the study river.
Nesting patterns associated with landforms can be provide insight into which flow stages (and their respective discharges) contribute to maintaining desirable aspects of channel topography such as riffle-pool sequences, or heterogeneity broadly.
Outputs
Directory Location: By default all nesting analysis outputs are saved @ GCS_analysis_outputs/nesting_analysis/.
Visual Outputs:
Static .png plots visualizing Ws, Zs, and C(Ws, Zs) series where each flow stage is stacked in a single plot, i.e., Ws_nesting_gcs_plots.png (see example below).
Interactive HTML “Sankey Diagrams” visualizing cross-section landform transitions between each increasing flow stage. The code outputs one with ‘normal’ cross-sections included, and another focusing only on non-normal GCS landforms (see the static example below).
Tabular Ouputs:
- A .csv table with Chi-squared significance levels associated for all landform transitions between flow stages. landform_transitions_chi_square.csv.
Note that the Chi-squared expected frequency parameter is defined by the relative landform abundance at the higher flow stage for a given transition.
For example, take a river where 50% of cross-sections that are nozzles (narrow and shallow) at a 0.7ft flow stage transition into wide bars (wide and shallow) at a 1.2ft flow stage. If wide bar cross-sections only represent 20% of the higher flow stage, we can say that nozzles preferentially transition into wide bars if significant at the p < 0.05 level.
- A .csv table with results from a T-Test comparing the the relative width (Ws) values of cross-sections with Zs < -0.5 and Zs > 0.5 at the next higher flow stage.
The idea here is to see whether high and low elevation cross-sections have some preferential relationship with channel geometry at the next-higher flow stage.
Output: preferential_nesting_ttest.csv.

