Geospatial tools for
NC PFAS Testing Network

Helena Mitasova

V. Petras,     B. Baines,     S. Reckling,     W. Reckling,     A. Petrasova    

       

Outline

  • PFAS Testing Network geospatial data overview
  • Drinking water sources: water systems, sources and sampling sites
  • Sampling routing tool v.net.tripplanner
  • Rapid viewing of GPS data with attributes
  • PFAS Testing Network on-line mapping tool prototype

  • Geospatial support: Groundwater seepage through streams, agreggated landfill and wastewater data

  • Future work: DEQ data, new sampling round, new analytics data layers, tools

PFAS Testing Network data management

PFAS Testing Network Data Hub: RENCI infrastructure

  • RENCI: integrated database with all collected data and related workflows
  • see PFAST website for the type of data collected
  • NCSU CGA: support geospatial data acquisition, processing, visualization

                 

PFAST Network geospatial data

Team 1. Public drinking water sources: 403 sites (NCSU, Duke)

PFAST Network geospatial data

Team 2. Chemours area: Groundwater seepage in streams, 24 samples (NCSU)

PFAST Network geospatial data

Team 4: Atmospheric monitoring sites: 11 sites (UNC-CH, UNC-W)

PFAST Network geospatial data

Team 5a: Landfills, waste water treatment plants: 20+ sites (NCSU)

PFAST Network geospatial data

Additional geospatial data collected:

  • Team 3: residential tap water (Duke)
  • Team 5b: fish, aligators (NCSU)
  • DEQ ~1000 private wells
  • Others ?

Integration of all PFAS data in RENCI data hub should provide a big picture of PFAS spatial and temporal distribution in North Carolina

Drinking water sources

  • Municipial surface water sources: all sampled
  • Multiple sources in some districts

Drinking water sources

  • Groundwater wells: only one per system sampled
  • Challenges indentifying the sampled well - GPS was crucial

Drinking water sources

  • All public groundwater and surface drinking water sources
  • Includes water sources serving small number of people

Drinking water sampled sites

  • All public groundwater and surface drinking water sources
  • 403 largest sources out of 8411 are sampled

Georeferencing drinking water samples

Several GPS coordinates may be associated with each drinking source
  • DEQ location (surface intake - blue, groundwater system - red)
  • sampled water source / intake / well - green
  • sampled tap

From spreadsheet to database

Duke integrates sample analysis results for Team 1 in SQLite database - each site has number of attributes, including the PFAS concentrations and GPS coordinates

Unique ID for each sample is critical for data management

Samples are matched with DEQ PWSWS data for additional variables: adress, contacts, well gw production, integration with GIS layers for further analysis

Surface water and groundwater sites by well yield and depth

Geospatial Tools

  • Sampling campaign planning tool
  • On-line viewer for point data
  • NC PFAST Data Viewer prototype

Sampling campaign planning tool

Scheduling and routing software tool to support optimized sampling of many sites: v.net.tripplanner implemented in GRASS GIS

  • travel from home to multiple sites over multiple trips
  • optimize number of trips needed at average travel speed
  • define maximum time per one trip, time spent at each place
  • schedule can be re-optimized after additional sites are added

Sampling campaign planning tool

Example routes 21 through 25 for second campaign

planned speed 45mi/h, actual 39mi/h; time spent at site planned 30 min, actual 21 min

On-line viewing tool for point data with attributes

Preview your spreadsheet data and create maps with open source kepler.gl
  • name the columns with GPS coordinates as name_latitude, name_longitude
  • select only "clean" data (avoid text with commas, NaN, empty cells, etc.)
  • export your sheet as csv
  • drop your csv file into kepler.gl (limit 250 Mb)

Color and size points by attributes

Convert to raster or heatmap

NC PFAST viewer prototype

WebGIS application based on PostGIS database, QGIS and custom tools

Developed as MGIST capstone project by Blake Baines

   

NC PFAST viewer prototype: display data

Display sampled sites colored by selected atribute (e.g. concentration)

Plan: make color scheme and symbology consistent with DEQ maps

NC PFAST viewer prototype: graph concentrations

Query a site, graph distribution of PFAS concentrations, display service area

Plan: make graph consistent with the reports generated by team 1 and RENCI data hub

NC PFAST viewer prototype: query data

Query selected site, show concentrations, associated service area and its attributes: population served, production capacity

NC PFAS viewer development

  • Integrate with RENCI data hub and DEQ data
  • Derive and add relevant data sets to support analysis and modeling:
    • contributing areas for each sampling site
    • map potential sources in contributing areas
    • map history of landuse since 1974 and facilities

Data aggregation

Team 5a Final Report: Turning graphs into maps - Mapping aggregated release for USGS Hydrologic Units 6

Future

  • Geospatial Analytics: derive relevant geospatial layers, explore spatial and temporal relationships using geostatistical modeling and machine learning to assist with understanding of sources, transport and fate
  • Geospatial Simulations: scenarios, decision support, where to go from here
  • Exploratory scientific visualization and public outreach data communication

       

Thank You!

NCSU GeoForAll Lab