Efficient stream gauging networks

This repository serves two functions:

1. Provides a code workflow and example analysis to illustrate how network coverage analysis was performed for the State Water Resources Control Board's 2022 Stream Gaging Plan. Lucy Andrews and Dr. Theodore Grantham supported the computational work underlying the plan. The plan was created in response to California Senate Bill 19.

2. Publish code used in a manuscript published in Nature Sustainability.

Title: Designing efficient stream gauging networks for twenty-first century water management

Abstract: Stream gauging stations provide critical information to water managers, but biases and gaps in gauge placement, compounded by gauge deactivations, limit our ability to track river flows and address global water challenges. Here we introduce a novel approach for identifying gauging sites that efficiently fill gaps in coverage within river networks while also addressing water management priorities, including reservoir operations, biodiversity conservation, and hydroclimatic monitoring. Applying this approach to gauges in California, USA, we found significant gaps in the stream monitoring network. Hypothetically reconfiguring gauges to locations that maximize coverage and representation of management objectives highlights the current network’s biases. Through the strategic reactivation and placement of additional gauges, we demonstrate how stream gauging networks can be designed to support sustainable water management.

Repository Structure

The code workflow is structured for scripts to be run in sequence. A user should run the file compile_analysis.Rmd, which does the following:

1. Sets up the environment
2. Downloads, imports, and cleans data that represent the stream network, stream gages, and management objectives (gaging dams, critical biodiversity, reference-quality streams, and natural communities commonly associated with groundwater)
3. Associates stream network segments with attributes relevant to management objectives
4. Analyzes network coverage provided by gages, both extant and hypothetical
5. Evaluates the coverage that the current gage network provides of management objectives
6. Runs a greedy set cover algorithm on potential gage locations to identify the most efficient set of gages that would cover the entire network for two scenarios (network expansion and network reconfiguration)
7. Ranks efficient gages by the coverage of management objectives that gages would provide.
8. Summarizes and visualizes results
9. Creates example figures that illustrate methods

.
└── stream_gage_final  
    ├── scripts  
        ├── 01_set_up  
            ├── 01_load_packages.R  
            ├── 02_set_options_parameters.R
            ├── 03_create_directories.R  
            ├── 04_build_functions.R
            ├── 05_build_base_maps.R
            ├── X_build_draft_functions.R
        ├── 02_clean_data  
            ├── 01_import_clean_nhd_wbd.R
            ├── 02_import_clean_ace.R
            ├── 03_import_clean_nccag.R
            ├── 04_import_clean_ref_streams.R
            ├── 05_import_clean_nid.R
            ├── 06_add_management_attributes.R
        ├── 03_analyze_networks  
            ├── 01_analyze_full_network.R
            ├── 02_reshape_network_analysis.R
            ├── 03_run_set_cover_expansion.R
            ├── 04_run_set_cover_all.R
            ├── 05_import_clean_nhd_wbd.R
            ├── 06_analyze_reconfigure_coverage.R
        ├── 04_create_output  
            ├── 01_create_text_output.R
            ├── 02_create_gaged_expansion_network_figures.R
            ├── 03_create_reconfig_network_figures.R
            ├── 04_create_comparison_figures.R
            ├── 05_create_manuscript_text_figures.R
            ├── 06_create_additional_manuscript_figures.R
            ├── 07_create_methods_example.R
        ├── compile_analysis.Rmd                            # key compilation script
    ├── data  
        ├── metadata
            ├── ace_aquatic_biodiversity_factsheet.pdf 
            ├── example_result.png
            ├── methods_flowchart.png
            ├── nccag_metadata_link.txt
            ├── nhdplus_v2_user_guide.pdf
        ├── raw_data                                        # created via set-up script
        ├── processed_data                                  # created via set-up script
    ├── output                                              # created via set-up script
        ├── figures                                         # created via set-up script
    ├── packages_unpublished  
        ├── StrmAnlyzeR_forked.zip  
    ├── .gitignore  
    ├── LICENSE  
    ├── README.md  

End Notes

Different entities spell "gauge" differently. Publishers tend to prefer "gauge", while American water management agencies prefer "gage". We use "gage" in our scripts (since they were originally conceived to support State Water Resources Control Board activities) but "gauge" in our writing. Apologies for the slight confusion this may create!