paper/paper.md

In IMMM-SFA/teleconnect: Geospatial Analytics for Multisector Urban Teleconnections

title: 'gamut: A Geospatial R Package to Analyze Multisectoral Urban Teleconnections' tags: - R - Multisector Dynamics - Water - Energy - Land - Urban - Geospatial date: "30 September 2021" output: html_document: df_print: paged authors: - name: Kristian D. Nelson orcid: 0000-0002-6745-167X affiliation: 1 - name: Sean W. Turner orcid: 0000-0003-4400-9800 affiliation: 1 - name: Chris R. Vernon orcid: 0000-0002-3406-6214 affiliation: 1 - name: Jennie S. Rice orcid: 0000-0002-7833-9456 affiliation: 1 bibliography: paper.bib affiliations: - name: Pacific Northwest National Laboratory, Richland, Washington, USA index: 1

Summary

Most cities in the United States withdraw surface water to meet public water supply needs. The lands on which this water is generated are often developed for human activities—such as agriculture, mining, and industry—that may compete for water resources or contaminate water supplies. Cities are thereby connected to other sectors through their water supply catchments. These connections are an example of a multisectoral urban teleconnection, or an interdependency to a geographically disparate region from a source region where events in one (e.g., land use changes) often impact the other [@Seto:2012]. This term was brought about to bring greater understanding of the connections between urbanization and land use changes [@Seto:2012]. The Geospatial Analytics for Multisectoral Urban Teleconnections (gamut) package provides national-scale information on these urban teleconnections for 235 cities by combining land use data with hydrological analysis to characterize and quantify urban source watershed human interactions across the conterminous United States (Figure 1).

{ width=95% }

The gamut package computes dozens of city-level metrics that inform on the geographical nature of surface water supply catchments and the presence, intensity, and impact of human activities in those catchments. The package cycles through a 3-step process. First, it connects cities to their drinking water resources by using the Urban Water Blueprint dataset [@McDonald:2014]. This dataset is combined with an enhancement dataset of source contribution estimates, river flow statistics, and high resolution runoff data[@Nelson:2021]. After linking the cities to the watersheds, the second step is using the watershed boundary as a mask for the input geospatial layers. Additionally, gamut relies heavily on the use of the st_intersection function from the sf package to intersect features and watersheds, which joins the information from the geospatial layers. These layers encompass a wide variety of watershed characteristics including land use, land cover, power generation, infrastructure, hydrological data (irrigation, stream flow, runoff), and population data. Similar to other multisector packages, gamut produces new data from a combination of multiple datasets, which are then used to create new statistics. Non-geospatial layers in the form of data tables are joined through city names and municipal IDs. The input layers used in this package have been combined into an open-source dataset and can be accessed here [@Nelson2:2021]. The last step in the process is the creation of output metrics, which is done through the calculation of statistics based on the masked input data for each city.

Creating the links between cities and watershed characteristics enables the gamut package to calculate numerous metrics that may be used for multiple types of city-level multisector dynamics research. Metrics reported by gamut fall into four main categories: geographical characteristics of watersheds (e.g., climate zones, land area, distance from city, hydrology), potential water contamination concentrations (nonpoint and point), withdrawal/consumption of water from other sectors, and presence/intensity of multisectoral land uses. Table 1 shows the metrics that are created and includes descriptions and units. An R vignette is provided to help users to get started with gamut and may be accessed here.

Table 1: Metrics reported in gamut

| Metric Name | Description | Units | | :------------------------------------ | :-------------------------------------------------------------------------------------- | :-------------------- | | city_population | The population of the city being analyzed | people | | n_watersheds | Number of watersheds that city uses to source drinking water | watersheds | | n_other_cities | Number of other cities pulling off the same watersheds | cities | | dependent_city_pop | Total population of people dependent on that city’s watersheds | people | | watershed_area_sqkm | Combined area of all the source watersheds of a city | square kilometers | | storage_BCM | Combined storage capacity of all the city catchments | billion cubic meters | | yield_BCM | Combined yield capacity of all the city catchments | billion cubic meters | | irr_cons_BCM | Combined water consumption that is used for irrigation with the watersheds | billion cubic meters | | n_climate_zones | Number of climate zones that the source watersheds cover | zones | | n_hydro_plants | Number of hydroelectric power plants operating within the source watersheds | plants | | n_thermal_plants | Number of thermal power plants operating within the source watersheds | plants | | n_fac_agcrop | Number of agricultural crop facilities within the source watersheds | facilities | | n_fac_aglivestock | Number of agicultural livestock facilities within the source watersheds | facilities | | n_fac_cnsmnf | Number of construction and manufacturing facilities within the source watersheds | facilities | | n_fac_mining | Number of mining facilities within the source watersheds | facilities | | n_fac_oilgas | Number of oil and gas facilities within the source watersheds | facilities | | n_fac_total | Total number of facilities operating within the source watersheds | facilities | | hydro_gen_MWh | Combined hydroelectric generation from all the facilities within the source watersheds | megawatt-hours | | thermal_gen_MWh | Combined thermal generation from all the facilities within the source watersheds | megawatt-hours | | thermal_cons_BCM | Combined water consumption that is used for thermal generation | billion cubic meters | | thermal_with_BCM | Combined water withdrawal for thermal generation | billion cubic meters | | n_utilities | Number of electric utilities within the source watersheds | utilities | | n_ba | Number of balancing authorities within the source watersheds | balancing authorities | | n_crop_classes | Total number of different types of crops within the source watersheds | crops | | cropland_fraction | Fraction of land that is used for crops | fraction | | developed_fraction | Fraction of land that is developed | fraction | | ag_runoff_max | Agricultural runoff as proportion of total runoff (worst-case watershed) | fraction | | ag_runoff_av_exgw | Agricultural runoff as proportion of total runoff in supply (exc. groundwater) | fraction | | ag_runoff_av | Agricultural runoff as proportion of total runoff in supply (inc. groundwater) | fraction | | dev_runof_max | Urban runoff as proportion of total runoff (worst-case watershed) | fraction | | dev_runof_av_exgw | Urban runoff as proportion of total runoff in supply (exc. groundwater) | fraction | | dev_runof_av | Urban runoff as proportion of total runoff in supply (inc. groundwater) | fraction | | np_runoff_max | Max amount of non-point source runoff within the source watersheds | fraction | | np_runoff_av_exgw | Nonpoint Proportion of Potentially Contaminated Supply (PPCS) (exc. groundwater) | fraction | | np_runoff_av_ exgw_unweighted | Nonpoint supply contamination averaged across watersheds | fraction | | np_runoff_av | Nonpoint Proportion of Potentially Contaminated Supply (PPCS) | fraction | | n_economic_sectors | Total number of different economic sectors within the source watersheds | sectors | | max_withdr_dis_km | Maximum distance between a city’s intake points | kilometers | | avg_withdr_dis_km | Average distance between a city’s intake points | kilometers | | n_treatment_plants | Total number of waste water treatment plants operating within the source watersheds | plants | | watershed_pop | Total number of people living within the source watershed boundaries | people | | pop_cons_m3sec | Combined water consumption from the source watersheds that is used for people | m3/sec | | av_fl_sur_conc_pct | Point PPCS (surface water only, based on flow) | % | | av_fl_sur_ conc_pct_unweighted | Point PPCS (surface water only, based on flow, not weighted by source importance) | % | | av_ro_sur_conc_pct | Point PPCS (surface water only, based on runoff) | % | | av_fl_all_conc_pct | Point PPCS (based on flow) | % | | av_ro_all_conc_pct | Point PPCS (based on runoff) | % | | av_fl_max_conc_pct | Point PPCS (based on flow, worst-case catchment only) | % | | av_ro_max_conc_pct | Point PPCS (based on runoff, worst-case catchment only) | % | | surface_contribution_pct | Proportion of total average supply made up from surface water | % | | importance_of_worst_ watershed_pct | Proportion of total average supply made up from most heavily contamined watershed | % | |

Statement of Need

Multisector Dynamics (MSD) research is the study of the co-evolution of human and natural systems. This research requires infrastructure expansion and land use scenarios, resource demand projections, and multisectoral modeling to capture the impacts of trends and shocks on human systems. The gamut package offers new data that meet a number of MSD needs. The package may be used to infer possible water resources expansion strategies for major cities in the United States. For example, cities found to be heavily exposed to potential contamination may be more likely to seek alternative means of supply (e.g., water transfers) or invest in water reuse facilities. In a study by @Rice:2013 which looked at de facto wastewater reuse across the US, it was found that there had been an increase in wastewater concentrations in drinking water treatment plants from a 1980 EPA report, especially at low flow conditions. The gamut package has the ability to look at wastewater discharge and average flow to find these concentrations at a much larger scale, showing that this package could be useful in studies like this in the future.

In addition to water contamination analysis, the gamut package has the ability to reveal which source watersheds are heavily protected by receiving cites. This information can inform land use and energy expansion scenarios applied in MSD research, for example by preventing significant expansion of human developments in protected source watersheds. gamut may also be used in large-scale hydrological modeling to correctly assign urban water demands to specific intakes. Whether research is being done on water scarcity, water pollution, or urbanization effects, the gamut package provides useful data that can brings greater understanding of anthropogenic impacts on urban source watersheds.

The gamut package is open source and may be downloaded using the devtools package with the code below [@Wickham:2020]. Further instructions on package download can be found in the documentation.

install.packages("devtools")
library(devtools)
devtools::install_github('IMMM-SFA/gamut')
library(gamut)

Dependencies

gamut relies on functionality from the following R packages: clisymbols [@Csardi2:2017], crayon [@Csardi:2017], dplyr [@Henry2:2020], dams [@Goteti:2020], exactextractr [@Baston:2020], foreign [@R-Core-Team:2020], geosphere [@Hijmans:2019], ggplot2 [@Wickham:2016], lwgeom [@Pebesma:2020], magrittr [@Bache:2014], purrr [@Henry:2020], raster [@Hijmans:2020], readxl [@Wickham:2019], reservoir [@Turner:2016], rgdal [@Bivand2:2020], rgeos [@Bivand:2020], sf [@Pebesma:2018], sp [@Bivand:2013], spex [@Sumner:2020], stringr [@Wickham2:2019], tibble [@Muller:2020], tidyr [@Wickham2:2020], vroom [@Hester:2021], testthat [@Wickham:2011], rmarkdown [@Xie:2018], knitr [@Xie:2014].

Acknowledgements

This research was supported by the U.S. Department of Energy, Office of Science, as part of research in MultisectorDynamics, Earth and Environmental System Modeling Program.

References

IMMM-SFA/teleconnect documentation built on Oct. 22, 2021, 2 p.m.