Generate Geographic Regional Summaries of Fishery Data (Version 1)

Creates geospatial representations of fishery metrics by aggregating landing site data to regional levels along the Zanzibar coast. This simplified version uses direct administrative district mappings instead of spatial proximity calculations, making it more efficient for cases where district-to-region relationships are already established.

Usage

create_geos_v1(monthly_summaries_dat = NULL, pars = NULL)

Arguments

monthly_summaries_dat

A data frame containing monthly fishery metrics by district. Required columns:

• district: Character, name of the landing site district

• date: Date, month of the summary

• mean_cpue_day: Numeric, mean catch per unit effort per day

• mean_rpue_day: Numeric, mean revenue per unit effort per day

• mean_price_kg: Numeric, mean price per kilogram

pars

Configuration parameters list containing:

• storage$google$key: Cloud storage provider key

• storage$google$options_coasts: Cloud storage options for coastal data

Value

NULL (invisible). The function uploads data to cloud storage as side effects:

• Parquet file: "zanzibar_monthly_summaries_map" containing regional summaries

• GeoJSON file: "ZAN_regions_version.geojson" containing regional boundaries

Details

The function performs the following operations:

1. District-Region Mapping: Retrieves pre-defined administrative mappings between districts and regions from site metadata

2. Data Harmonization: Standardizes district names (e.g., "Chake chake" to "Chake Chake") to ensure proper joining

3. Regional Aggregation: Calculates monthly summary statistics for each region by aggregating district-level data

4. GeoJSON Export: Combines regional polygon geometries with summary statistics and exports to cloud storage

Key Differences from create_geos():

• Uses administrative mappings (ADM column) instead of spatial distance calculations

• More efficient as it avoids complex spatial operations

• Relies on pre-established district-region relationships in metadata

• Handles district name inconsistencies automatically

Calculated Regional Metrics (using median values across districts in each region):

• Mean CPUE (Catch Per Unit Effort, kg per fisher per day)

• Mean RPUE (Revenue Per Unit Effort, TZS per fisher per day)

• Mean Price per kg of catch

Output Format:

• Regional summaries are exported as Parquet files to cloud storage

• GeoJSON file containing regional boundaries is uploaded separately

• Both files use standardized naming conventions with version information

Note

Dependencies:

• Requires the sf package for reading and writing spatial data

• Requires "ZAN_coast_regions.geojson" file in the package inst/ directory

• Uses get_metadata() function to retrieve district-region mappings

• Uses add_version() to append version information to filenames

• Uses upload_parquet_to_cloud() and upload_cloud_file() for cloud storage

Data Processing Notes:

• District names are case-corrected (specifically "Chake chake" → "Chake Chake")

• Regions are converted to lowercase in the output for consistency

• Dates are formatted in ISO 8601 format (YYYY-MM-DDTHH:MM:SSZ)

• All outputs include "country" = "zanzibar" for multi-country compatibility

See also

• create_geos() for the spatial proximity-based version

• export_wf_data() which calls this function as part of the export workflow

• get_metadata() for retrieving site and administrative information

Examples

if (FALSE) { # \dontrun{
# Load configuration
pars <- read_config()

# Get monthly summaries data
monthly_data <- download_parquet_from_cloud(
  prefix = "wf_monthly_summaries",
  provider = pars$storage$google$key,
  options = pars$storage$google$options
)

# Create regional geospatial summaries
create_geos_v1(
  monthly_summaries_dat = monthly_data,
  pars = pars
)
} # }