kidwellj/mapping_environmental_action
1
0
Fork 0
mirror of https://github.com/kidwellj/mapping_environmental_action.git synced 2024-10-31 23:42:20 +00:00
Code Issues Packages Projects Releases Wiki Activity

updated readme

Browse source
This commit is contained in:
Jeremy Kidwell 2020-03-26 18:08:49 +00:00
parent ec980c423d
commit 61f39ecd34
4 changed files with 342 additions and 965 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

62
README.md
View file

@ -1,10 +1,12 @@
# Mapping Environmental Action
This code is associated with a scholarly paper "Mapping Environmental Action" (currently unpublished). In the paper, I draw on original data gathered from my research with Eco-groups in Scotland (2013-2017) in order to do comparative geospatial analysis of the coincidence of these groups with a number of standard demographics. You can read (an unpublished version of) the paper at (http://mapenvcom.jeremykidwell.info/mapping_draft.html).
## Why Reproducible Research? ##
If you're new to github and reproducible research, welcome! It's nice to have you here. Github is ordinarily a place where software developers working on open source software projects deposit their code as they write software collaboratively. However, in recent years a number of scholarly researchers, especially people working on research which involves a digital component (including me!) have begun to deposit their papers in these same software repositories. The idea here is that you can download all of the source-code and data used in this paper alongside the actual text, run it yourself and ["reproduce" the results](http://kbroman.org/steps2rr/). This can serve as a useful safeguard, a layer of research transparency, and a cool teaching tool for other persons interested in doing similar work.
If you're new to github and reproducible research, welcome! It's nice to have you here. Github is ordinarily a place where software developers working on open source software projects deposit their code as they write software collaboratively. However, in recent years a number of scholarly researchers, especially people working on research that involves a digital component (including me!) have begun to deposit their papers in these same software repositories. The idea is that you can download all of the source-code and data used in this paper alongside the actual text, run it yourself and ["reproduce" the results](http://kbroman.org/steps2rr/). This can serve as a useful safeguard, a layer of research transparency, and a cool teaching tool for other persons interested in doing similar work. Particularly when, as is the case in subject areas that are only just starting to get involved in the digital humanities, like religious studies, there is a dearth of work of this nature, it can be helpful to have examples of practice which can be reused, or at least used as an example.
Eschewing proprietary, expensive and unreliable software like Microsoft Word, I write in a combination of two languages: (1) [Markdown](https://en.wikipedia.org/wiki/Markdown) which is intended to be as close as possible to plain text while still allowing for things like boldfaced type, headings and footnotes; and (2) a programming language called [R](https://en.wikipedia.org/wiki/R_(programming_language)) to do all the data analysis. R is an object oriented language which was specifically designed for statistical analysis. It's also great fun to tinker with. As you look through this paper, you'll see that R code is integrated into the text of the document. This is indicated by a series of three backticks (```). There is a formal specification now at a mature stage of development, which is RMarkdown. You can read semi-official specification [for this here](https://bookdown.org/yihui/rmarkdown/pdf-document.html).
Eschewing proprietary, expensive and unreliable software like Microsoft Word, I write in a combination of two languages: (1) [Markdown](https://en.wikipedia.org/wiki/Markdown) which is intended to be as close as possible to plain text while still allowing for things like boldfaced type, headings and footnotes; and (2) a programming language called [R](https://en.wikipedia.org/wiki/R_(programming_language)) to do all the data analysis. R is an object oriented language that was specifically designed for statistical analysis. It's also great fun to tinker with. As you look through this paper, you'll see that R code is integrated into the text of the document. This is indicated by a series of three backticks (```). There is a formal specification now at a mature stage of development, which is RMarkdown. You can read semi-official specification [for this here](https://bookdown.org/yihui/rmarkdown/pdf-document.html).
To read a bit more on these things and start on your own path towards plain text reproducible research, I highly recommend:
- Karl Broman's guide, "[Initial Steps Toward Reproducible Research](http://kbroman.org/steps2rr/)"
@ -12,21 +14,22 @@ To read a bit more on these things and start on your own path towards plain text
The other advantage of putting this paper here is that readers and reviewers can suggest changes and point out errors in the document. To do this, I recommend that you create a github issue by clicking on the green "New issue" button [here](https://github.com/kidwellj/mapping_environmental_action/issues). If you must, you can also send me emails. More stuff about me [can be found here](http://jeremykidwell.info).
To skip ahead and start reading the actual paper, click on [`mapping_draft.rmd`](https://github.com/kidwellj/mapping_environmental_action/blob/master/mapping_draft.Rmd) above.
To skip ahead and start reading the actual paper in raw format, click on [`mapping_draft.rmd`](https://github.com/kidwellj/mapping_environmental_action/blob/master/mapping_draft.Rmd) above. If you were looking for the article (without code) you can also find a working draft here: (http://mapenvcom.jeremykidwell.info/mapping_draft.html).
Now for...
## The quick technical version ##
## The technical version ##
This repository contains the code and writing towards a (working draft of a) scholarly paper which presents my analysis of the geospatial footprint of eco-groups in the UK. This is based on research I have been conducting since 2013 and which is ongoing. The paper is written in R Markdown and for the most part, I'm using the conventions outlined by Kieran Healy [here](https://kieranhealy.org/blog/archives/2014/01/23/plain-text/) and is best viewed (I think) in [R Studio](https://www.rstudio.com) though it will be reasonably comprehensible to anyone using a Markdown editor. If I'm not working in RStudio, I'm probably in Sublime text, FYI. Co-authors and collaborators take note, generally, I use [Hadley Wickham's venerable R Style Guide](http://adv-r.had.co.nz/Style.html).
This repository contains the code and writing towards a (working draft of a) scholarly paper that presents my analysis of the geospatial footprint of eco-groups in the UK. This is based on research I have been conducting since 2013 and that is ongoing. The paper is written in R Markdown and for the most part, I'm using the conventions outlined by Kieran Healy [here](https://kieranhealy.org/blog/archives/2014/01/23/plain-text/) and is best viewed (I think) in [R Studio](https://www.rstudio.com) though it will be reasonably comprehensible to anyone using a Markdown editor. If I'm not working in RStudio, I'm probably in Sublime text, FYI. Co-authors and collaborators take note, generally, I use [Hadley Wickham's venerable R Style Guide](http://adv-r.had.co.nz/Style.html).
I'd be extremely happy if someone found errors, or imagined a more efficient means of analysis and either reported them as an issue on this github repository or sent me an email.
The actual article is in `mapping_draft.Rmd` and can be compiled using knitr (assuming you have R installed as well as required packages) using the script provided `knit_it_html.sh`
The actual article is in `mapping_draft.Rmd` and can be compiled using knitr (assuming you have R installed as well as required packages) using the `Makefile` provided.
Note: actual execution may take over an hour, as calls to `st_buffer` and `st_within` under `wilderness_data_prep` are computationally intensive. To compile more briskly, I recommend you comment out this final section and knit the markdown/html files.
Note: actual execution may take over an hour, as calls to `st_buffer` and `st_within` under `wilderness_data_prep` are computationally intensive. To compile more briskly, I recommend you comment out this final section and knit the markdown/html files. I have been relying on the University of Birmingham supercomputing cluster for execution, which has resulted in a parallel version of this script `mapping_draft-hpc_optimised.Rmd`. The latter will only run on the BlueBEAR cluster at UOB, though other scholars may want to consult this script to get a sense of how geospatial operations can be parallelised for more efficient execution.
Paths in this folder are used mostly for R processing. Towards this end folders have the following significance:
Paths in this folder are used mostly for R processing. I'm using a "project" oriented workflow, on which you can read more [in a blog by Jenny Bryan here](https://www.tidyverse.org/blog/2017/12/workflow-vs-script/). This uses the R package [here](https://cran.r-project.org/web/packages/here/index.html).
Towards this end folders have the following significance:
- `data` contains datasets used for analysis.
- `derived_data` contains files which represent modified forms of files in the above path.
@ -34,3 +37,46 @@ Paths in this folder are used mostly for R processing. Towards this end folders
- `cache` isn't included in github but is usually used for working files
Note: none of the contents of the above are included in the github repository unless they are unavailable from an external repository.
## And, a few notes for the data scientists
Over the course of this research project (since 2013, really), the state of geospatial tools for datascience in R (and python) has shifted and the increased attention and resources that have been brought to bear on geospatial has resulted in a dramatic improvement in the quality and precision of tools available, particularly the development of SimpleFeatures and the [sf](https://github.com/r-spatial/sf) and [tmap](https://github.com/mtennekes/tmap) packages for R. Ggplot2 is awesome, but starts to creak quickly when you push it in more creative geospatial directions. There are also inefficiencies with data handling in some of the older packages (such as sp) that aren't apparent until you start working with large datasets. Underlying data formats have been shifting quite a lot as well, from csvt and [very problematic and proprietary ESRI shapefiles](http://switchfromshapefile.org/) to geojson/topojson and [Geopackage](http://switchfromshapefile.org/#geopackage) formats. The result of this has a need to completely rewrite this script mid-way through the research process. I've left some of the messy bits in with as comprehensive comments as possible to give a sense of things, but there remain some bits which are accidentally messy.
There are a few aspects of this code which are novel or were difficult that I'm proud of, which I hope may be useful and on ehich I'd especially value
- The use of sf() and tmap()
- The creation of vignettes for visualisations
- The level of reproducibility
- Optimisation of intensive geospatial operations for htpc and parallel computing
# Prerequisites for reproducing this codebase
I've tried to follow best practices in setting up this script for reproducibility, but given some of the choices I've had to make computationally (e.g. running some operations in PostGIS) some setup is required before execution will be successful.
These steps are:
1. Acquire a working installation of R (and RStudio). I have produced a Docker container that replicates the environment I have used to execute this script that is probably the easiest way to complete this task.
2. Set up a working Postgres database with PostGIS extensions installed. The script will download necessary data and load it into your database if it is not already in place.
3. Install platform appropriate prerequisites for the R odbc() package, see here: [https://github.com/r-dbi/odbc#installation]
4. Configure a local `config.yml` file with the following information (used to connect to your PostGIS database):
```
default:
datawarehouse:
driver: 'Postgres'
server: 'change.to.yourserver.com'
uid: 'change-to-your-username'
pwd: 'change-to-your-password'
port: 5432
database: 'database-name'
```
5. Clone or download the code from this repository
6. Set up a proper R/RStudio working environment. I use the `renv` package to manage working environment, which takes snapshots and stores them to `renv.lock`. If you run `renv::restore()` in R after loading this code, it will install necessary libraries at proper versions.
7. Nearly all of the data used in this study is open, with one exception, that of the Ordnance Survey PointX data product. This is available to most UK academics via the EDINA service, so the user will need to manually download this data and place it in the `/data/` directory.
# Contributing
Please note that this project is released with a [Contributor Code of Conduct](CODE_OF_CONDUCT.md). By contributing to this project, you agree to abide by its terms.
# License
The content of this research paper are licensed under the [Creative Commons Attribution-ShareAlike 4.0 International Public License](https://creativecommons.org/licenses/by-sa/4.0/legalcode), and the underlying source code used to generate the paper is licensed under the [GNU AGPLv3](https://www.gnu.org/licenses/agpl-3.0.en.html) license. Underlying datasets designed as part of this research have their own licenses that are specified in their respective repositories.

2
data/scotland_admin_2011pop.csv
View file

@ -1,4 +1,4 @@
Council areas,2011_pop,CODE
council areas,2011_pop,code
Aberdeen City,"222,793",S12000033
Aberdeenshire,"252,973",S12000034
Angus,"115,978",S12000041

1 Council areas council areas 2011_pop CODE code
2 Aberdeen City 222,793 S12000033
3 Aberdeenshire 252,973 S12000034
4 Angus 115,978 S12000041

469
mapping_draft-hpc_optimised.Rmd
View file

@ -36,48 +36,65 @@ output:
---
```{r setup, include=FALSE}
require(knitr)
require(kableExtra)
# Note, this script has been written largely with RStudio on MacOS, but is compiled on
# an hpc cluster which runs Linux, so some tweaks below ensure smooth execution
# in both environments. Also, as above, the script is meant to output to both PDF and
# html_document knitr formats.
require(knitr) # used to knit RMarkdown format script into working documents in various formats
require(kableExtra) # used for markdown table formatting compatible with knitr
# note: some features of the below line are specific to html/pdf format and will need to be adapted pre-compile until dual outputs are working (see https://github.com/kidwellj/mapping_environmental_action/issues/2)
knitr::opts_chunk$set(fig.path='figures/', warning=FALSE, echo=FALSE, message=FALSE, dpi=300, fig.width=7)
# TODO: consider implementing knitcitations - https://github.com/cboettig/knitcitations
# TODO: fix simultaneous output towards PDF, see here: https://stackoverflow.com/questions/23621012/display-and-save-the-plot-simultaneously-in-r-rstudio
```
```{r load_packages, message=FALSE, warning=FALSE, include=FALSE}
## Default repo
# setwd("/Users/jeremy/gits/mapping_environmental_action")
# setwd("/Users/kidwellj/OneDrive\ -\ bham.ac.uk/writing/201708_mapping_environmental_action")
# Set up working machine independent working directory and environment
require(here) # used to keep working directory organised and portable
require(renv) # used to set up environment
require(usethis) # integrates git support for rStudio
# Set repository to be new standard, e.g. cloud server.
# This will avoid a dialogue box if packages are to be installed for below on first run.
# For smooth execution on command-line knitr - this will avoid a dialogue box
# if packages are to be installed for below on first run
local({r <- getOption("repos")
r["CRAN"] <- "https://cloud.r-project.org"
options(repos=r)
})
# TODO: remove sp etc. once sf is fully implemented
# TODO: automatically test for packages below on given execution environment and run install.packages() as needed.
require(RCurl) # used for fetching reproducible datasets
require(sf) # new simplefeature data class, supercedes sp in many ways
require(sp) # needed for proj4string, deprecated by sf()
require(rgdal) # deprecated by sf()
require(GISTools) # deprecated by sf()
require(rgeos) # deprecated by sf()
require(maptools)
# See issue https://github.com/kidwellj/mapping_environmental_action/issues/3 for progress re: migration from sp()
# require(sp) # needed for proj4string, deprecated by sf()
# require(maptools)
require(ggplot2)
require(tmap) # using as an alternative to base r graphics and ggplot for geospatial plots
require(tmaptools) # for get_asp_ratio below
require(grid) # using for inset maps on tmap
require(broom) # required for tidying SPDF to data.frame for ggplot2
require(tidyr) # using for grouped bar plot
require(plyr)
require(dplyr)
# require(plyr) # already a dependency of knitr, remove?
# require(dplyr) # already a dependency of knitr, remove?
require(reshape2) # using for grouped bar plot
require(scales)
# require(sqldf) # using sqldf to filter before loading very large data sets
## Packages required for PostGIS database access
# Many thanks to Sébastien Rochette for documentation here: https://www.r-bloggers.com/interact-with-postgis-from-r/
# require(config) # used to access database connection credentials in config.yml file below
# library(DBI)
# library(odbc)
# library(RPostgres)
# library(rpostgis)
# library(dbplyr)
## Packages required for knitr output
## Packages used for features or issues relating to html_document knitr output
require(plotly) # allows for export of plots to dynamic web pages
require(gtable) # more powerful package for multi-plot layouts, not necessary for knitr
require(showtext) # for loading in fonts
require(extrafont) # font support
## Packages used for features or issues relating to pdf_document knitr format
# Note: implementation of fonts (currently commented out) is specific to pdf_document output
# require(showtext) # for loading in fonts
# require(extrafont) # font support
# Set up local workspace:
if (dir.exists("data") == FALSE) {
@ -90,18 +107,33 @@ if (dir.exists("derivedData") == FALSE) {
dir.create("derivedData")
}
# # Setup PostGIS database connection
# dw <- config::get("datawarehouse")
#
# con <- dbConnect(odbc::odbc(),
# driver = dw$driver,
# database = dw$database,
# uid = dw$uid,
# pwd = dw$pwd,
# host = dw$server,
# port = 5432
# )
# Define Coordinate Reference Systems (CRS) for later use
# Note: I've used British National Grid (27000) in this paper, but have found that
# it is falling out of use in many cases, so will be defaulting to WGS84 in future
# data-sets and papers.
# Working with EPSG codes for spatialfeature CRS given the usage of this approach with sf()
# for discussion related to this fix, see https://gis.stackexchange.com/q/313761/41474
# TODO: remove below as part of overall migration to sf()
# See issue https://github.com/kidwellj/mapping_environmental_action/issues/3 for progress re: migration from sp()
bng <- CRS("+init=epsg:27700")
wgs84 <- CRS("+init=epsg:4326")
# Configure fonts for plots below
## Configure fonts for plots below, commented out currently because of incompatibilities
## Loading Google fonts (http://www.google.com/fonts)
# Note: implementation of fonts (currently commented out) is specific to pdf_document output
# font_add_google("Merriweather", "merriweather")
# The following will load in system fonts (uncomment and run as needed on first execution)
# font_import(pattern="[A/a]rial", prompt=FALSE)
@ -116,16 +148,24 @@ Until recently, environmentalism has been treated by governments and environment
```{r load_ecs_data, message=FALSE, warning=FALSE}
# read in Eco-Congregation Scotland data and-------------------
# ...turn it into a SpatialPointsDataFrame---------------------
# TODO: upload ECS-GIS-Locations_3.0.csv to zenodo repository, i.e.
ecs <- read.csv("data/ECS-GIS-Locations_3.0.csv", comment.char="#")
# unnecessary with advent of sf (above)
coordinates(ecs) <- c("X", "Y")
# Modified to use EPSG code directly 27 Feb 2019
proj4string(ecs) <- bng
ecs_sf <- st_as_sf(ecs, coords = c("X", "Y"), crs=paste0("+init=epsg:",27700))
# TODO: update below to match new dataset once it has been uploaded to zenodo
# if (file.exists("data/ECS-GIS-Locations_3.0.geojson") == FALSE) {
# download.file("https://____.zip",
# destfile = "data/____.zip")
# unzip("data/____.zip", exdir = "data")
# }
# Note, use of paste0 here relates to fix noted above.
# for discussion related to this approach, see https://gis.stackexchange.com/q/313761/41474
# read in Eco-Congregation Scotland data and-------------------
# ...turn it into a SimpleFeature---------------------
ecs <- st_read("data/ECS-GIS-Locations_3.0.geojson") %>% st_transform(paste0("+init=epsg:",27700))
# Write data to PostGIS database for later analysis, currently commented out for later integration
# st_write(ecs, dsn = con, layer = "ecs",
# overwrite = FALSE, append = FALSE)
```
There are `r length(ecs)` eco-congregations in Scotland. By some measurements, particularly in terms of individual sites and possibly also with regards to volunteers, this makes Eco-Congregation Scotland one of the largest environmental third-sector groups in Scotland.[^159141043]
There are `r length(ecs$name)` eco-congregations in Scotland. By some measurements, particularly in terms of individual sites and possibly also with regards to volunteers, this makes Eco-Congregation Scotland one of the largest environmental third-sector groups in Scotland.[^159141043]
In seeking to conduct GIS and statistical analysis of ECS, it is important to note that there some ways in which these sites are statistically opaque. Our research conducted through interviews at a sampling of sites and analysis of a variety of documents suggests that there is a high level of diversity both in terms of the number of those participating in environmental action and the types of action underway at specific sites. Work at a particular site can also ebb and flow over the course of time. Of course, as research into other forms of activism and secular environmental NGOs has shown, this is no different from any other third sector volunteer group. Variability is a regular feature of groups involved in activism and/or environmental concern.
@ -141,7 +181,7 @@ For the sake of comparison, we also measured the geographical footprint of two o
# Technical Background
Analysis was conducted using QGIS 2.8 and R `r getRversion()`, and data-sets were generated in CSV format.[^15541313] To begin with, I assembled a data set consisting of x and y coordinates for each congregation in Scotland and collated this against a variety of other specific data. Coordinates were checked by matching UK postcodes of individual congregations against geo-referencing data in the Office for National Statistics postcode database. In certain instances a single "congregation" is actually a series of sites which have joined together under one administrative unit. In these cases, each site was treated as a separate data point if worship was held at that site at least once a month, but all joined sites shared a single unique identifier. As noted above, two other datasets were generated for the sake of comparative analysis.[^177171536] These included one similar Environmental Non-Governmental Organisation (ENGO) in Scotland (1) Transition Scotland (which includes Scotland Communities Climate Action Network);[^15541342] and another community-based NGO, Scottish Community Development Trusts.[^158261232] As this report will detail, these three overlap in certain instances both literally and in terms of their aims, but each also has a separate identity and footprint in Scotland. Finally, in order to normalise data, we utilised the PointX POI dataset which maintains a complete database of Places of Worship in Scotland.[^15541614]
Analysis was conducted using QGIS 3.12 and R `r getRversion()`, and data-sets were generated in CSV, geopackage, and geojson format.[^15541313] To begin with, I assembled a data set consisting of x and y coordinates for each congregation in Scotland and collated this against a variety of other specific data. Coordinates were checked by matching UK postcodes of individual congregations against geo-referencing data in the Office for National Statistics postcode database. In certain instances a single "congregation" is actually a series of sites which have joined together under one administrative unit. In these cases, each site was treated as a separate data point if worship was held at that site at least once a month, but all joined sites shared a single unique identifier. As noted above, two other datasets were generated for the sake of comparative analysis.[^177171536] These included one similar Environmental Non-Governmental Organisation (ENGO) in Scotland (1) Transition Scotland (which includes Scotland Communities Climate Action Network);[^15541342] and another community-based NGO, Scottish Community Development Trusts.[^158261232] As this report will detail, these three overlap in certain instances both literally and in terms of their aims, but each also has a separate identity and footprint in Scotland. Finally, in order to normalise data, we utilised the PointX POI dataset which maintains a complete database of Places of Worship in Scotland.[^15541614]
# Background and History of Eco-Congregation Scotland
@ -149,17 +189,16 @@ Eco-Congregation Scotland began a year before the official launch of Eco-Congreg
```{r calculate_ecs_by_year, message=FALSE, warning=FALSE}
# Tidy up date fields and convert to date data type
ecs$registration <- as.Date(ecs$registration, "%Y-%m-%d")
# TODO: Fix issues here with R complaining that "character string is not in a std...
# ecs$award1 <- as.Date(ecs$award1)
# ecs$award2 <- as.Date(ecs$award2)
# ecs$award3 <- as.Date(ecs$award3)
# ecs$award4 <- as.Date(ecs$award4)
ecs$registration <- as.Date(ecs$registration, "%d/%m/%Y")
ecs$award1 <- as.Date(ecs$award1, "%d/%m/%Y")
ecs$award2 <- as.Date(ecs$award2, "%d/%m/%Y")
ecs$award3 <- as.Date(ecs$award3, "%d/%m/%Y")
ecs$award4 <- as.Date(ecs$award4, "%d/%m/%Y")
# TODO: add "R" to command in paragraph below once this is resolved, do search for all non 'r instances
ecs_complete_cases <- ecs[complete.cases(ecs$year_begun),]
```
The programme launched officially in 2001 at Dunblane Cathedral in Stirling and by 2005 the project had `r length(ecs_complete_cases[ecs_complete_cases$year_begun < 2006, ])` congregations registered to be a part of the programme and 25 which had completed the curriculum successfully and received an Eco-Congregation award. By 2011, the number of registrations had tripled to `r length(ecs_complete_cases[ecs_complete_cases$year_begun < 2012, ])` and the number of awarded congregations had quadrupled to `sum(ecs$award1 < "01/01/2012", na.rm=TRUE)`. This process of taking registrations and using a tiered award or recognition scheme is common to many voluntary organisations. The ECS curriculum was developed in part by consulting the Eco-Congregation England and Wales materials which had been released just a year earlier in 1999, though it has been subsequently revised, particularly with a major redesign in 2010. In the USA, a number of similar groups take a similar approach including Earth Ministry (earthministry.org) and Green Faith (greenfaith.org).
The programme launched officially in 2001 at Dunblane Cathedral in Stirling and by 2005 the project had `r length(which(as.numeric(as.character(ecs$year_begun)) < 2006))` congregations registered to be a part of the programme and 25 which had completed the curriculum successfully and received an Eco-Congregation award. By 2011, the number of registrations had tripled to `r length(which(as.numeric(as.character(ecs$year_begun)) < 2012))` and the number of awarded congregations had quadrupled to `r length(which((ecs$award1 < "2012/01/01")))`. This process of taking registrations and using a tiered award or recognition scheme is common to many voluntary organisations. The ECS curriculum was developed in part by consulting the Eco-Congregation England and Wales materials which had been released just a year earlier in 1999, though it has been subsequently revised, particularly with a major redesign in 2010. In the USA, a number of similar groups take a similar approach including Earth Ministry (earthministry.org) and Green Faith (greenfaith.org).[^footnote1]
In the case of Eco-Congregation Scotland, congregations are invited to begin by "registering" their interest in the programme by completing a basic one-sided form. The next step requires the completion of an award application, which includes a facilitated curriculum called a "church check-up" and after an application is submitted, the site is visited and assessed by third-party volunteer assessors. Sites are invited to complete additional applications for further awards which are incremental (as is the application process). Transition communities, at least in the period reflected on their map, go through a similar process (though this does not involve the use of a supplied curriculum) by which they are marked first as "interested," become "active" and then gain "official" status.[^1554162]
@ -167,16 +206,15 @@ In the case of Eco-Congregation Scotland, congregations are invited to begin by
```{r import_admin_data, message=FALSE, warning=FALSE, include=FALSE}
# read in polygon for Scottish admin boundaries
# TODO: upload bundle of admin data to new zenodo repository and alter below to use new URLs
# TODO: need to remove readOGR below once st_read is confirmed to be working as sf
if (file.exists("data/scotland_ca_2010.shp") == FALSE) {
download.file("https://borders.ukdataservice.ac.uk/ukborders/easy_download/prebuilt/shape/Scotland_ca_2010.zip",
destfile = "data/Scotland_ca_2010.zip")
unzip("data/Scotland_ca_2010.zip", exdir = "data")
}
admin_lev1 <- readOGR("./data", "scotland_ca_2010")
admin_lev1_sf <- st_read("data/scotland_ca_2010.shp") %>% st_transform(paste0("+init=epsg:",27700))
admin_lev1 <- st_read("data/scotland_ca_2010.shp") %>% st_transform(paste0("+init=epsg:",27700))
# Write to SQL database
# st_write(admin_lev1, dsn = con, layer = "admin_lev1",
# overwrite = FALSE, append = FALSE)
# read in polygon for intermediate admin boundary layers
if (file.exists("data/scotland_parlcon_2011.shp") == FALSE) {
@ -184,12 +222,9 @@ download.file("http://census.edina.ac.uk/ukborders/easy_download/prebuilt/shape/
destfile = "data/Scotland_parlcon_2011.zip")
unzip("data/Scotland_parlcon_2011.zip", exdir = "data")
}
admin_lev2 <- readOGR("./data", "scotland_parlcon_2011")
admin_lev2_sf <- st_read("data/scotland_parlcon_2011.shp") %>% st_transform(paste0("+init=epsg:",27700))
# Set CRS using epsg code on spdf for symmetry with datasets below
proj4string(admin_lev1) <- bng
proj4string(admin_lev2) <- bng
admin_lev2 <- st_read("data/scotland_parlcon_2011.shp") %>% st_transform(paste0("+init=epsg:",27700))
st_write(admin_lev2, dsn = con, layer = "admin_lev2",
overwrite = FALSE, append = FALSE)
# Generate new sf shape using bounding box for central belt for map insets below
# Note: coordinates use BNG as CRS (EPSG: 27700)
@ -210,99 +245,117 @@ scotland_ratio<- get_asp_ratio(scotland)
```{r import_groups_data, message=FALSE, warning=FALSE, include=FALSE}
# read in Transition Towns data and turn it into a SpatialPointsDataFrame
transition_wgs <- read.csv(text=getURL("https://zenodo.org/record/165519/files/SCCAN_1.4.csv"))
coordinates(transition_wgs) <- c("X", "Y")
proj4string(transition_wgs) <- CRS("+init=epsg:4326")
transition <- spTransform(transition_wgs, bng)
transition_sf <- st_as_sf(transition, coords = c("X", "Y"), crs=27700)
# read in all_churches data (data set generated by Jeremy Kidwell to replace PointX data used from Ordnance Survey)
# TODO: need to remove all data points which are outside BNG area to resolve error
# also need to make symmetrical with ECS denominations, add Methodist
# churches, remove nazarene and salvation army
# Original approach using sp()
# transition_wgs <- read.csv(text=getURL("https://zenodo.org/record/165519/files/SCCAN_1.4.csv"))
# coordinates(transition_wgs) <- c("X", "Y")
# proj4string(transition_wgs) <- CRS("+init=epsg:4326")
# transition <- spTransform(transition_wgs, bng)
# transition_sf <- st_as_sf(transition, coords = c("X", "Y"), crs=27700)
# churches_all <- read.csv("data/all_churches_0.9.csv")
# churches_all_clean<-churches_all[complete.cases(churches_all),]
# churches_all_null<-churches_all[!complete.cases(churches_all),]
# coordinates(churches_all) <- c("X", "Y")
# proj4string(churches_all) <- proj4string(admin_lev1)
transition <- st_read("data/transition-scotland_2.3.geojson") %>% st_transform(paste0("+init=epsg:",27700))
# read in pointX data and turn it into a SpatialPointsDataFrame
pow_pointX <- read.csv("./data/poi_2015_12_scot06340459.csv", sep="|")
coordinates(pow_pointX) <- c("feature_easting", "feature_northing")
# TODO: need to alter to draw from wgs84 or bng defined in preamble above
proj4string(pow_pointX) <- CRS("+init=epsg:27700")
pow_pointX_sf <- st_as_sf(pow_pointX, coords = c("X", "Y"), crs=27700)
# TODO to add code here to parse out raw pointx file (or OS local?)
# read in Scottish Community Dev. trust data and turn it into a SpatialPointsDataFrame
dtas <- read.csv("data/community-dev-trusts-2.6.csv")
coordinates(dtas) <- c("X", "Y")
proj4string(dtas) <- CRS("+init=epsg:27700")
dtas_sf <- st_as_sf(dtas, coords = c("X", "Y"), crs=27700)
# Original approach using sp()
# pow_pointX <- read.csv("./data/poi_2015_12_scot06340459.csv", sep="|")
# coordinates(pow_pointX) <- c("feature_easting", "feature_northing")
# # TODO: need to alter to draw from wgs84 or bng defined in preamble above
# proj4string(pow_pointX) <- CRS("+init=epsg:27700")
# pow_pointX_sf <- st_as_sf(pow_pointX, coords = c("X", "Y"), crs=27700)
# read in permaculture data and turn it into a SpatialPointsDataFrame
permaculture <- read.csv("data/permaculture_scot-0.8.csv")
coordinates(permaculture) <- c("X", "Y")
proj4string(permaculture) <- CRS("+init=epsg:27700")
permaculture_sf <- st_as_sf(permaculture, coords = c("X", "Y"), crs=27700)
pow_pointX <- st_read("data/pointx_201512_scotland_pow.geojson") %>% st_transform(paste0("+init=epsg:",27700))
## read in Scottish Community Dev. trust data and turn it into a SpatialPointsDataFrame
# removing deprecated use of sp() and csv
# dtas <- read.csv("data/community-dev-trusts-2.6.csv")
# coordinates(dtas) <- c("X", "Y")
# proj4string(dtas) <- CRS("+init=epsg:27700")
# dtas_sf <- st_as_sf(dtas, coords = c("X", "Y"), crs=27700)
dtas <- st_read("data/community-dev-trusts-2.6.geojson") %>% st_transform(paste0("+init=epsg:",27700))
## read in permaculture data and turn it into a SpatialPointsDataFrame
# removing deprecated use of sp() and csv
# permaculture <- read.csv("data/permaculture_scot-0.8.csv")
# coordinates(permaculture) <- c("X", "Y")
# proj4string(permaculture) <- CRS("+init=epsg:27700")
# permaculture_sf <- st_as_sf(permaculture, coords = c("X", "Y"), crs=27700)
permaculture <- st_read("data/permaculture_scot-0.8.geojson") %>% st_transform(paste0("+init=epsg:",27700))
# subset point datasets for inset maps below
ecs_sf_centralbelt <- st_intersection(ecs_sf, centralbelt_region)
dtas_sf_centralbelt <- st_intersection(dtas_sf, centralbelt_region)
transition_sf_centralbelt <- st_intersection(transition_sf, centralbelt_region)
pow_sf_centralbelt <- st_intersection(pow_pointX_sf, centralbelt_region)
permaculture_sf_centralbelt <- st_intersection(permaculture_sf, centralbelt_region)
# commenting out deprecated naming convention
# ecs_sf_centralbelt <- st_intersection(ecs_sf, centralbelt_region)
# dtas_sf_centralbelt <- st_intersection(dtas_sf, centralbelt_region)
# transition_sf_centralbelt <- st_intersection(transition_sf, centralbelt_region)
# pow_sf_centralbelt <- st_intersection(pow_pointX_sf, centralbelt_region)
# permaculture_sf_centralbelt <- st_intersection(permaculture_sf, centralbelt_region)
ecs_centralbelt <- st_intersection(ecs, centralbelt_region)
dtas_centralbelt <- st_intersection(dtas, centralbelt_region)
transition_centralbelt <- st_intersection(transition, centralbelt_region)
pow_centralbelt <- st_intersection(pow_pointX, centralbelt_region)
permaculture_centralbelt <- st_intersection(permaculture, centralbelt_region)
```
```{r process_admin_data}
# Augment existing dataframes to run calculations and add columns with point counts per polygon,
# percentages, and normalising data.
# This code will generate a table of frequencies for each spatialpointsdataframe in admin
# Generate a table of frequencies for each set of points in admin_lev1
# calculate count of ECS for fields in admin and provide percentages
# JK Note: need to convert from poly.counts, which uses sp() to st_covers() which uses sf() - cf. https://stackoverflow.com/questions/45314094/equivalent-of-poly-counts-to-count-lat-long-pairs-falling-inside-of-polygons-w#45337050
admin_lev1$ecs_count <- poly.counts(ecs,admin_lev1)
# Thanks to commenter for tip on sf() approach here at https://stackoverflow.com/questions/45314094/equivalent-of-poly-counts-to-count-lat-long-pairs-falling-inside-of-polygons-w#45337050
admin_lev1$ecs_count <- lengths(st_covers(admin_lev1, ecs))
admin_lev1$ecs_percent<- prop.table(admin_lev1$ecs_count)
# # Test new approach
# library(sp)
# library(GISTools)
# admin_lev1$ecs_count_sp <- poly.counts(ecs,admin_lev1)
# admin_lev1$ecs_count_sp - admin_lev1$ecs_count
# admin_lev1$ecs_percent <- prop.table(admin_lev1$ecs_count)
# calculate count of places of worship in PointX db for fields in admin and provide percentages
admin_lev1$pow_count <- poly.counts(pow_pointX,admin_lev1)
# TODO ingest data here from OS Open Map Local dataset (rather than non-open PointX dataset; need to filter and upload to zenodo)
admin_lev1$pow_count <- lengths(st_covers(admin_lev1, pow_pointX))
admin_lev1$pow_percent<- prop.table(admin_lev1$pow_count)
# calculate count of Transition for fields in admin and provide percentages
admin_lev1$transition_count <- poly.counts(transition,admin_lev1)
admin_lev1$transition_count <- lengths(st_covers(admin_lev1, transition))
admin_lev1$transition_percent<- prop.table(admin_lev1$transition_count)
# calculate count of dtas for fields in admin and provide percentages
admin_lev1$dtas_count <- poly.counts(dtas,admin_lev1)
admin_lev1$dtas_count <- lengths(st_covers(admin_lev1, dtas))
admin_lev1$dtas_percent<- prop.table(admin_lev1$dtas_count)
# calculate count of permaculture for fields in admin and provide percentages
admin_lev1$permaculture_count <- poly.counts(permaculture,admin_lev1)
admin_lev1$permaculture_count <- lengths(st_covers(admin_lev1, permaculture))
admin_lev1$permaculture_percent<- prop.table(admin_lev1$permaculture_count)
# run totals for intermediate boundaries level 2
# This code will generate a table of frequencies for each spatialpointsdataframe in admin_lev2
# calculate count of ECS for fields in admin_lev2 and provide percentages
admin_lev2$ecs_count <- poly.counts(ecs,admin_lev2)
admin_lev2$ecs_count <- lengths(st_covers(admin_lev2, ecs))
admin_lev2$ecs_percent<- prop.table(admin_lev2$ecs_count)
# calculate count of places of worship in PointX db for fields in admin_lev2 and provide percentages
admin_lev2$pow_count <- poly.counts(pow_pointX,admin_lev2)
admin_lev2$pow_count <- lengths(st_covers(admin_lev2, pow_pointX))
admin_lev2$pow_percent<- prop.table(admin_lev2$pow_count)
# calculate count of Transition for fields in admin_lev2 and provide percentages
admin_lev2$transition_count <- poly.counts(transition,admin_lev2)
admin_lev2$transition_count <- lengths(st_covers(admin_lev2, transition))
admin_lev2$transition_percent<- prop.table(admin_lev2$transition_count)
# calculate count of dtas for fields in admin_lev2 and provide percentages
admin_lev2$dtas_count <- poly.counts(dtas,admin_lev2)
admin_lev2$dtas_count <- lengths(st_covers(admin_lev2, dtas))
admin_lev2$dtas_percent<- prop.table(admin_lev2$dtas_count)
# calculate count of permaculture for fields in admin_lev2 and provide percentages
admin_lev2$permaculture_count <- poly.counts(permaculture,admin_lev2)
admin_lev2$permaculture_count <- lengths(st_covers(admin_lev2, permaculture))
admin_lev2$permaculture_percent<- prop.table(admin_lev2$permaculture_count)
# calculate count of ECS for fields in admin_lev2_sf
# original approach - calculate count of ECS for fields in admin_lev2_sf
# TODO: for future migration to sf throughout, remove above content and swap out references.
admin_lev2_sf$ecs_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(ecs_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$pow_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(pow_pointX_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$transition_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(transition_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$dtas_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(dtas_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$permaculture_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(permaculture_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# admin_lev2_sf$ecs_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(ecs_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# admin_lev2_sf$pow_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(pow_pointX_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# admin_lev2_sf$transition_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(transition_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# admin_lev2_sf$dtas_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(dtas_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# admin_lev2_sf$permaculture_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(permaculture_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# Import csv with population data for each level of administrative subdivision and join to spatialdataframe
@ -313,9 +366,13 @@ admin_lev2_sf$permaculture_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(pe
# TODO - consider adapting to use ONS mid-year statistics: https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/populationestimatesforukenglandandwalesscotlandandnorthernireland
JK revising done to here 2020-Mar-11
# Load population statistics for normalising data by population and merge with admin_lev1
admin_lev1_pop <- read.csv("./data/scotland_admin_2011pop.csv")
# TODO: sf() merge doesn't support by.x etc. operations, need to adapt
admin_lev1 <- merge(x=admin_lev1, y=admin_lev1_pop, by.x = "code", by.y = "CODE")
admin_lev1_merged <- merge(admin_lev1, admin_lev1_pop)
# Convert number stored as string with a comma to an integer
admin_lev1$X2011_pop <- as.numeric(gsub(",", "", admin_lev1@data$X2011_pop))
# Calculate counts as percentages of total for normalising below
@ -351,13 +408,12 @@ admin_lev2$ecs_count_pownorm_scaled <- admin_lev2$ecs_count_pownorm*(sum(admin_l
Perhaps the first important question to ask of these groups is, where are they? I calculated the spread of eco-congregations and transition groups across each of the 32 council areas in Scotland. Every council area in Scotland has at least one eco-congregation or transition group). The most are located in `r as.character(admin_lev1$NAME_2[which.max(admin_lev1$ecs_count)])`, with `r max(admin_lev1$ecs_count)`, whereas the mean among all the 32 council areas is `r mean(admin_lev1$ecs_count)`, with a median of `r median(admin_lev1$ecs_count)`, standard deviation of `r sd(admin_lev1$ecs_count)`, and interquartile range of `r IQR(admin_lev1$ecs_count)`. The following choropleth maps show the relative concentration of eco-congregations (indicated by yellow to red).
Though there are too few eco-congregations and transition groups for a numerically significant representation in any of the intermediate geographies, mapping the concentration of sites by agricultural parishes allows for a more granular visual and I include this for comparison sake. Note, for the sake of a more accurate visual communication, we have also marked out areas of Scotland that are uninhabited with hash marks on the map of agricultural parishes. (*TODO: this will be done in the final draft, once I get my image masking fixed!*).[^15571030]
Though there are too few eco-congregations and transition groups for a numerically significant representation in any of the intermediate geographies, mapping the concentration of sites by agricultural parishes allows for a more granular visual and I include this for comparison sake.[^15571030]
## Eco-Congregation Scotland groups shown by concentration in administrative regions (NUTS3)
```{r 01_admin_ecs_choropleth, fig.width=4, fig.cap="Figure 1"}
# Note: for more information on EU administrative levels, see here: https://ec.europa.eu/eurostat/web/nuts/national-structures-eu
# TODO: clip choropleth polygons to buildings shapefile (possble superceded by pverlay on lev2)
# Draw initial choropleth map of ECS concentration (using tmap and sf below by default)
# Revising re: CRS inset maps complete to here
@ -370,7 +426,7 @@ tm_shape(admin_lev2) +
# also consider scaling text size using area # quick plot example:
# qtm(World, fill = "income_grp", text = "iso_a3", text.size = "AREA") # use "World$" to see the two attributes: income_grp and iso_a3, text.size= area: text is sized increasingly with coutry area size.
# tm_text("name", size=.2, shadow=TRUE, bg.color="white", bg.alpha=.25) +
# tm_shape(ecs_sf) +
# tm_shape(ecs) +
# tm_dots("red", size = .05, alpha = .4) +
# tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
@ -451,7 +507,7 @@ admin_lev1$ecs_dtas_factor <- ((admin_lev1$ecs_percent - admin_lev1$dtas_percent
We can use this data to normalise our figures regarding Eco-Congregation Scotland communities and this draws the presence in Edinburgh of ECS communities into even sharper relief, as Edinburgh, though ranked second in terms of population and fifth in terms of places of worship, ranks first for the presence of all ECS congregations and awarded ECS congregations. However, taking population as the basis for normalisation first, we find that Edinburgh is far from the most prominent outlier. In trying to communicate this difference for a lay-audience, we have chosen to list this difference as a multiplier (i.e. there are 2.x times as many congregations as their share of population and an average figure of congregations might allow for) as this conveys the difference in a straight-forward way. Outliers where the disparity between their relative share of the total ECS footprint and their relative share of population is different by a positive ratio of more than double include the Orkney Islands (3.7 times more eco-congregations than their expected average share based on population), Argyll and Bute (`admin_lev1[CODE=S12000023]$ecs_pop_factor` 4.2x), Stirling (2.76x), and Perthshire and Kinross (2.18x). Interestingly, there are no outliers whose relative share of the total footprint of ECS is double or more in the negative direction (see Appendix A chart for full numbers).
Turning to the total of `r length(pow_pointX)` "places of worship" in Scotland, we find a slightly different picture of the relative concentration of Eco-Congregations in Scotland. In this case, the outliers are
Turning to the total of `r length(pow_pointX)` "places of worship" in Scotland, we find a slightly different picture of the relative concentration of Eco-Congregations in Scotland. In this case, the outliers are *TODO: add code here!*
Whereas our initial measurements indicated a prominent lead for Edinburgh, by normalising our data in this way we can highlight the stronger-than-expected presence of several others that might otherwise escape notice because they lie in a region with significantly lower population or numerically less places of worship. Taking the PointX data on "places of worship" in Scotland, we find a less dramatic picture, but also a slightly different one. The positive outliers include East Renfrewshire (3.4x) Edinburgh (2.9x), Stirling (2.2), West Lothian (1.9x) and Aberdeen (1.5x). Again, negative outliers are far less dramatic, with only Midlothian possessing a ratio of more than 100% negative difference from the number of "places of worship" at 1.5x *fewer*.
@ -497,7 +553,7 @@ admin_lev2_scotland_ecs_plot <-
tm_borders(alpha=.5, lwd=0.1) +
tm_shape(admin_lev1) +
tm_borders(lwd=0.6) +
tm_shape(ecs_sf) +
tm_shape(ecs) +
tm_dots("red", size = .02, alpha = .2) +
tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
@ -517,12 +573,12 @@ admin_lev2_scotland_ecs_plot <-
# Build smaller central belt plot for inset:
admin_lev2_centralbelt <- st_crop(admin_lev2_sf, centralbelt_region)
admin_lev2_centralbelt <- st_crop(admin_lev2, centralbelt_region)
admin_lev2_centralbelt_ecs_plot <-
tm_shape(admin_lev2_centralbelt) +
tm_fill(col = "ecs_count", palette = "Oranges", n = 5) +
tm_shape(ecs_sf_centralbelt) + tm_dots("red", size = .05, alpha = .4) +
tm_shape(ecs_centralbelt) + tm_dots("red", size = .05, alpha = .4) +
tm_legend(show=FALSE)
# Stitch together maps using grid()
@ -543,7 +599,7 @@ tm_shape(admin_lev2) +
tm_borders(alpha=.5, lwd=0.1) +
tm_shape(admin_lev1) +
tm_borders(lwd=0.6) +
tm_shape(transition_sf) +
tm_shape(transition) +
tm_dots("red", size = .02, alpha = .2) +
tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
@ -566,7 +622,7 @@ tm_shape(admin_lev2) +
tm_borders(alpha=.5, lwd=0.1) +
tm_shape(admin_lev1) +
tm_borders(lwd=0.6) +
tm_shape(dtas_sf) +
tm_shape(dtas) +
tm_dots("red", size = .02, alpha = .2) +
tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
@ -590,7 +646,7 @@ tm_shape(admin_lev2) +
tm_borders(alpha=.5, lwd=0.1) +
tm_shape(admin_lev1) +
tm_borders(lwd=0.6) +
tm_shape(permaculture_sf) +
tm_shape(permaculture) +
tm_dots("red", size = .02, alpha = .2) +
tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
@ -656,28 +712,25 @@ download.file("http://sedsh127.sedsh.gov.uk/Atom_data/ScotGov/ZippedShapefiles/S
destfile = "data/SG_UrbanRural_2016.zip")
unzip("data/SG_UrbanRural_2016.zip", exdir = "data")
}
# Todo: remove sp datasets when sf revisions are complete. Currently running in parallel
urbanrural <- readOGR("./data", "SG_UrbanRural_2016")
proj4string(urbanrural) <- bng
urbanrural_sf <- st_read("data/SG_UrbanRural_2016.shp") %>% st_transform(paste0("+init=epsg:",27700))
urbanrural_sf_simplified <- st_simplify(urbanrural_sf)
urbanrural <- st_read("data/SG_UrbanRural_2016.shp") %>% st_transform(paste0("+init=epsg:",27700))
urbanrural_simplified <- st_simplify(urbanrural)
# TODO: worth considering uploading data to zenodo for long-term reproducibility as ScotGov shuffles this stuff around periodically breaking URLs
# This code will generate a table of frequencies for each spatialpointsdataframe in urbanrural
# calculate count of ECS for fields in urbanrural
urbanrural$ecs_count <- poly.counts(ecs,urbanrural)
urbanrural$ecs_count <- lengths(st_covers(urbanrural, ecs))
urbanrural$ecs_percent<- prop.table(urbanrural$ecs_count)
# calculate count of places of worship in PointX db for fields in urbanrural and provide percentages
urbanrural$pow_count <- poly.counts(pow_pointX,urbanrural)
urbanrural$pow_count <- lengths(st_covers(urbanrural, pow_pointX))
urbanrural$pow_percent<- prop.table(urbanrural$pow_count)
# calculate count of Transition for fields in urbanrural
urbanrural$transition_count <- poly.counts(transition,urbanrural)
urbanrural$transition_count <- lengths(st_covers(urbanrural, transition))
urbanrural$transition_percent<- prop.table(urbanrural$transition_count)
# calculate count of dtas for fields in urbanrural
urbanrural$dtas_count <- poly.counts(dtas,urbanrural)
urbanrural$dtas_count <- lengths(st_covers(urbanrural, dtas))
urbanrural$dtas_percent<- prop.table(urbanrural$dtas_count)
# calculate count of permaculture for fields in urbanrural
urbanrural$permaculture_count <- poly.counts(permaculture,urbanrural)
urbanrural$permaculture_count <- lengths(st_covers(urbanrural, permaculture))
urbanrural$permaculture_percent<- prop.table(urbanrural$permaculture_count)
```
@ -717,17 +770,15 @@ ggplot(urbanrural_gathered,
```{r 08_urbanrural_ecs_chart_choropleth, message=FALSE, warning=FALSE, fig.width=4, fig.cap="Figure 8"}
# TODO: Clip shapes to buildings shapefile (use OSM or OS?), using st_difference
# TODO: Double check data licenses for tm_credits
# Generate code for inset map of central belt
# First build large plot using National level view
# TODO: add inner margin to layout to provide adequate spce for inset map at bottom
urbanrural_uk_ecs_choropleth_plot <- tm_shape(urbanrural_sf_simplified) +
urbanrural_uk_ecs_choropleth_plot <- tm_shape(urbanrural_simplified) +
tm_polygons(col = "UR8FOLD", palette = "BrBG", lwd=0.001, n=9,
title = "UrbanRural 8 Fold Scale") +
tm_shape(ecs_sf) +
tm_shape(ecs) +
tm_dots("red", size = .05, alpha = .2) +
tm_scale_bar(position = c("left", "bottom")) +
tm_style("gray") +
@ -748,12 +799,12 @@ urbanrural_uk_ecs_choropleth_plot <- tm_shape(urbanrural_sf_simplified) +
# Next build smaller central belt plot for inset:
urbanrural_sf_simplified_centralbelt <- st_crop(urbanrural_sf_simplified, centralbelt_region)
urbanrural_simplified_centralbelt <- st_crop(urbanrural_simplified, centralbelt_region)
urbanrural_centralbelt_ecs_choropleth_plot <-
tm_shape(urbanrural_sf_simplified_centralbelt) +
tm_shape(urbanrural_simplified_centralbelt) +
tm_polygons(col = "UR8FOLD", palette = "BrBG") +
tm_shape(ecs_sf_centralbelt) + tm_dots("red", size = .05, alpha = .4) +
tm_shape(ecs_centralbelt) + tm_dots("red", size = .05, alpha = .4) +
tm_legend(show=FALSE)
# Stitch together maps using grid()
@ -777,8 +828,8 @@ print(urbanrural_centralbelt_ecs_choropleth_plot, vp = vp_urbanrural_centralbelt
# TODO: change basemap
# tmap_mode("view")
# tm_shape(urbanrural_sf_simplified) + tm_polygons(col = "UR8FOLD", palette = "BrBG") +
# tm_shape(ecs_sf) +
# tm_shape(urbanrural_simplified) + tm_polygons(col = "UR8FOLD", palette = "BrBG") +
# tm_shape(ecs) +
# tm_dots("red", size = .05, alpha = .4, popup.vars = TRUE) +
# tm_view(alpha = 1, basemaps = "Esri.WorldTopoMap")
@ -971,11 +1022,11 @@ scenicareas_simplified <- st_simplify(scenicareas)
# Set symmetrical CRS for analysis below (inserted here in order to correct errors, may be deprecated later)
# st_crs(sssi) <- 27700
# st_crs(ecs_sf) <- 27700
# st_crs(pow_pointX_sf) <- 27700
# st_crs(dtas_sf) <- 27700
# st_crs(transition_sf) <- 27700
# st_crs(permaculture_sf) <- 27700
# st_crs(ecs) <- 27700
# st_crs(pow_pointX) <- 27700
# st_crs(dtas) <- 27700
# st_crs(transition) <- 27700
# st_crs(permaculture) <- 27700
# Define buffer and measure number of ECS groups within 0.5 miles of all SSSI
# CRS uses meters for units, so a buffer for 0.5 miles would use 805 meters)
@ -1026,46 +1077,55 @@ scenicareas_buf500 <- st_buffer(scenicareas_simplified, dist = 500)
# plot(lnd[ sel, ], col = "turquoise", add = TRUE) # add selected zones to map
# from https://gotellilab.github.io/Bio381/StudentPresentations/SpatialDataTutorial.html
ecs_sf_sssi <- st_within(ecs_sf, sssi_simplified)
ecs_sf_sssi50m <- st_within(ecs_sf, sssi_buf50)
ecs_sf_sssi500m <- st_within(ecs_sf, sssi_buf500)
ecs_sf_sssibeyond500m <- !(st_within(ecs_sf, sssi_buf500))
# TODO: integrate pre-calc here into calculations further down which are still recalculating these figures
ecs_sssi <- st_within(ecs, sssi_simplified)
ecs_sssi50m <- st_within(ecs, sssi_buf50)
ecs_sssi500m <- st_within(ecs, sssi_buf500)
ecs_sssibeyond500m <- !(st_within(ecs, sssi_buf500))
ecs_sf_wildland <- st_within(ecs_sf, wildland_simplified)
ecs_sf_wildland50m <- st_within(ecs_sf, wildland_buf50)
ecs_sf_wildland500m <- st_within(ecs_sf, wildland_buf500)
ecs_sf_wildlandbeyond500m <- !(st_within(ecs_sf, wildland_buf500))
ecs_wildland <- st_within(ecs, wildland_simplified)
ecs_wildland50m <- st_within(ecs, wildland_buf50)
ecs_wildland500m <- st_within(ecs, wildland_buf500)
ecs_wildlandbeyond500m <- !(st_within(ecs, wildland_buf500))
ecs_sf_forestinv <- st_within(ecs_sf, forestinv_simplified)
ecs_sf_forestinv50m <- st_within(ecs_sf, forestinv_buf50)
ecs_sf_forestinv500m <- st_within(ecs_sf, forestinv_buf500)
ecs_sf_forestinvbeyond500m <- !(st_within(ecs_sf, forestinv_buf500))
ecs_forestinv <- st_within(ecs, forestinv_simplified)
ecs_forestinv50m <- st_within(ecs, forestinv_buf50)
ecs_forestinv500m <- st_within(ecs, forestinv_buf500)
ecs_forestinvbeyond500m <- !(st_within(ecs, forestinv_buf500))
ecs_sf_scenicareas <- st_within(ecs_sf, scenicareas_simplified)
ecs_sf_scenicareas50m <- st_within(ecs_sf, scenicareas_buf50)
ecs_sf_scenicareas500m <- st_within(ecs_sf, scenicareas_buf500)
ecs_sf_scenicareasbeyond500m <- !(st_within(ecs_sf, scenicareas_buf500))
ecs_scenicareas <- st_within(ecs, scenicareas_simplified)
ecs_scenicareas50m <- st_within(ecs, scenicareas_buf50)
ecs_scenicareas500m <- st_within(ecs, scenicareas_buf500)
ecs_scenicareasbeyond500m <- !(st_within(ecs, scenicareas_buf500))
# TODO: implement more efficient code using do.call() function or sapply() as here https://stackoverflow.com/questions/3642535/creating-an-r-dataframe-row-by-row
# TODO: implement parallel computing to distribute execution of loopable calculations below
# See: https://nceas.github.io/oss-lessons/parallel-computing-in-r/parallel-computing-in-r.html
# Generate dataframe based on SSSI buffers
# Calculate incidence of ecs within SSSI and within buffers at 50/500m
ecs_sssi_row <- c(sum(apply(st_within(ecs_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, sssi_buf500, sparse=FALSE), 1, any)))
ecs_sssi_row <- c(sum(apply(st_within(ecs, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, sssi_buf500, sparse=FALSE), 1, any)))
pow_sssi_row <- c(sum(apply(st_within(pow_pointX_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, sssi_buf500, sparse=FALSE), 1, any)))
pow_sssi_row <- c(sum(apply(st_within(pow_pointX, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, sssi_buf500, sparse=FALSE), 1, any)))
dtas_sssi_row <- c(sum(apply(st_within(dtas_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, sssi_buf500, sparse=FALSE), 1, any)))
dtas_sssi_row <- c(sum(apply(st_within(dtas, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, sssi_buf500, sparse=FALSE), 1, any)))
transition_sssi_row <- c(sum(apply(st_within(transition_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, sssi_buf500, sparse=FALSE), 1, any)))
transition_sssi_row <- c(sum(apply(st_within(transition, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition, sssi_buf500, sparse=FALSE), 1, any)))
permaculture_sssi_row <- c(sum(apply(st_within(permaculture_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, sssi_buf500, sparse=FALSE), 1, any)))
permaculture_sssi_row <- c(sum(apply(st_within(permaculture, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, sssi_buf500, sparse=FALSE), 1, any)))
grocery_sssi_row <- c(sum(apply(st_within(poi_grocery, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, sssi_buf500, sparse=FALSE), 1, any)))
pubs_sssi_row <- c(sum(apply(st_within(poi_pubs, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, sssi_buf500, sparse=FALSE), 1, any)))
# Generate dataframe from rows based on counts
sssi_counts <- rbind(ecs_sssi_row, pow_sssi_row)
sssi_counts <- rbind(sssi_counts, dtas_sssi_row)
sssi_counts <- rbind(sssi_counts, transition_sssi_row)
sssi_counts <- rbind(sssi_counts, permaculture_sssi_row)
sssi_counts <- rbind(sssi_counts, grocery_sssi_row)
sssi_counts <- rbind(sssi_counts, pubs_sssi_row)
sssi_counts <- as.data.frame(sssi_counts)
colnames(sssi_counts) <- c("Within SSSIs", "...50m", "...500m")
@ -1076,11 +1136,15 @@ pow_sssi_row_pct <- pow_sssi_row/length(pow_pointX)
dtas_sssi_row_pct <- dtas_sssi_row/length(dtas)
transition_sssi_row_pct <- transition_sssi_row/length(transition)
permaculture_sssi_row_pct <- permaculture_sssi_row/length(permaculture)
grocery_sssi_row_pct <- grocery_sssi_row/length(poi_grocery)
pubs_sssi_row_pct <- pubs_sssi_row/length(poi_pubs)
sssi_counts_pct <- rbind(ecs_sssi_row_pct, pow_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, dtas_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, transition_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, permaculture_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, grocery_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, pubs_sssi_row_pct)
colnames(sssi_counts_pct) <- c("% Within SSSIs", "% within 50m", "% within 500m")
# Merge into larger dataframe
@ -1088,34 +1152,44 @@ sssi_counts_merged <- cbind(sssi_counts, sssi_counts_pct)
# Generate dataframe based on wildland buffers
ecs_wildland_row <- c(sum(apply(st_within(ecs_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, wildland_buf500, sparse=FALSE), 1, any)))
ecs_wildland_row <- c(sum(apply(st_within(ecs, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, wildland_buf500, sparse=FALSE), 1, any)))
pow_wildland_row <- c(sum(apply(st_within(pow_pointX_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, wildland_buf500, sparse=FALSE), 1, any)))
pow_wildland_row <- c(sum(apply(st_within(pow_pointX, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(ecs_wildland_row, pow_wildland_row)
dtas_wildland_row <- c(sum(apply(st_within(dtas_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, wildland_buf500, sparse=FALSE), 1, any)))
dtas_wildland_row <- c(sum(apply(st_within(dtas, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, dtas_wildland_row)
transition_wildland_row <- c(sum(apply(st_within(transition_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, wildland_buf500, sparse=FALSE), 1, any)))
transition_wildland_row <- c(sum(apply(st_within(transition, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, transition_wildland_row)
permaculture_wildland_row <- c(sum(apply(st_within(permaculture_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, wildland_buf500, sparse=FALSE), 1, any)))
permaculture_wildland_row <- c(sum(apply(st_within(permaculture, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, permaculture_wildland_row)
grocery_wildland_row <- c(sum(apply(st_within(poi_grocery, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, grocery_wildland_row)
pubs_wildland_row <- c(sum(apply(st_within(poi_pubs, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, pubs_wildland_row)
colnames(wildland_counts) <- c("Within Wildland Areas", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
ecs_wildland_row_pct <- ecs_wildland_row/length(ecs_sf)
ecs_wildland_row_pct <- ecs_wildland_row/length(ecs)
pow_wildland_row_pct <- pow_wildland_row/length(pow_pointX)
dtas_wildland_row_pct <- dtas_wildland_row/length(dtas)
transition_wildland_row_pct <- transition_wildland_row/length(transition)
permaculture_wildland_row_pct <- permaculture_wildland_row/length(permaculture)
grocery_wildland_row_pct <- grocery_wildland_row/length(poi_grocery)
pubs_wildland_row_pct <- pubs_wildland_row/length(poi_pubs)
wildland_counts_pct <- rbind(ecs_wildland_row_pct, pow_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, dtas_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, transition_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, permaculture_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, grocery_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, pubs_wildland_row_pct)
colnames(wildland_counts_pct) <- c("% Within wildlands", "% within 50m", "% within 500m")
# Merge into larger dataframe
@ -1123,33 +1197,44 @@ wildland_counts_merged <- cbind(wildland_counts, wildland_counts_pct)
# Generate dataframe based on forestinv buffers
ecs_forestinv_row <- c(sum(apply(st_within(ecs_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, forestinv_buf500, sparse=FALSE), 1, any)))
ecs_forestinv_row <- c(sum(apply(st_within(ecs, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, forestinv_buf500, sparse=FALSE), 1, any)))
pow_forestinv_row <- c(sum(apply(st_within(pow_pointX_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, forestinv_buf500, sparse=FALSE), 1, any)))
pow_forestinv_row <- c(sum(apply(st_within(pow_pointX, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(ecs_forestinv_row, pow_forestinv_row)
dtas_forestinv_row <- c(sum(apply(st_within(dtas_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, forestinv_buf500, sparse=FALSE), 1, any)))
dtas_forestinv_row <- c(sum(apply(st_within(dtas, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, dtas_forestinv_row)
transition_forestinv_row <- c(sum(apply(st_within(transition_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, forestinv_buf500, sparse=FALSE), 1, any)))
transition_forestinv_row <- c(sum(apply(st_within(transition, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, transition_forestinv_row)
permaculture_forestinv_row <- c(sum(apply(st_within(permaculture_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, forestinv_buf500, sparse=FALSE), 1, any)))
permaculture_forestinv_row <- c(sum(apply(st_within(permaculture, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, permaculture_forestinv_row)
grocery_forestinv_row <- c(sum(apply(st_within(poi_grocery, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, grocery_forestinv_row)
pubs_forestinv_row <- c(sum(apply(st_within(poi_pubs, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, pubs_forestinv_row)
colnames(forestinv_counts) <- c("Within Woodlands", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
ecs_forestinv_row_pct <- ecs_forestinv_row/length(ecs_sf)
# TODO: fix error generated by ecs_forestinv_row_pct using ecs as sf().
ecs_forestinv_row_pct <- ecs_forestinv_row/length(ecs)
pow_forestinv_row_pct <- pow_forestinv_row/length(pow_pointX)
dtas_forestinv_row_pct <- dtas_forestinv_row/length(dtas)
transition_forestinv_row_pct <- transition_forestinv_row/length(transition)
permaculture_forestinv_row_pct <- permaculture_forestinv_row/length(permaculture)
grocery_forestinv_row_pct <- grocery_forestinv_row/length(poi_grocery)
pubs_forestinv_row_pct <- pubs_forestinv_row/length(poi_pubs)
forestinv_counts_pct <- rbind(ecs_forestinv_row_pct, pow_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, dtas_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, transition_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, permaculture_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, grocery_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, pubs_forestinv_row_pct)
colnames(forestinv_counts_pct) <- c("% Within Woodlands", "% within 50m", "% within 500m")
# Merge into larger dataframe
@ -1157,34 +1242,44 @@ forestinv_counts_merged <- cbind(forestinv_counts, forestinv_counts_pct)
# Generate dataframe based on scenicareas buffers
ecs_scenicareas_row <- c(sum(apply(st_within(ecs_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
ecs_scenicareas_row <- c(sum(apply(st_within(ecs, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs, scenicareas_buf500, sparse=FALSE), 1, any)))
pow_scenicareas_row <- c(sum(apply(st_within(pow_pointX_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
pow_scenicareas_row <- c(sum(apply(st_within(pow_pointX, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(ecs_scenicareas_row, pow_scenicareas_row)
dtas_scenicareas_row <- c(sum(apply(st_within(dtas_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
dtas_scenicareas_row <- c(sum(apply(st_within(dtas, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, dtas_scenicareas_row)
transition_scenicareas_row <- c(sum(apply(st_within(transition_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
transition_scenicareas_row <- c(sum(apply(st_within(transition, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, transition_scenicareas_row)
permaculture_scenicareas_row <- c(sum(apply(st_within(permaculture_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
permaculture_scenicareas_row <- c(sum(apply(st_within(permaculture, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, permaculture_scenicareas_row)
grocery_scenicareas_row <- c(sum(apply(st_within(poi_grocery, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, grocery_scenicareas_row)
pubs_scenicareas_row <- c(sum(apply(st_within(poi_pubs, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, pubs_scenicareas_row)
colnames(scenicareas_counts) <- c("Within Scenic Areas", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
ecs_scenicareas_row_pct <- ecs_scenicareas_row/length(ecs_sf)
ecs_scenicareas_row_pct <- ecs_scenicareas_row/length(ecs)
pow_scenicareas_row_pct <- pow_scenicareas_row/length(pow_pointX)
dtas_scenicareas_row_pct <- dtas_scenicareas_row/length(dtas)
transition_scenicareas_row_pct <- transition_scenicareas_row/length(transition)
permaculture_scenicareas_row_pct <- permaculture_scenicareas_row/length(permaculture)
grocery_scenicareas_row_pct <- grocery_scenicareas_row/length(poi_grocery)
pubs_scenicareas_row_pct <- pubs_scenicareas_row/length(poi_pubs)
scenicareas_counts_pct <- rbind(ecs_scenicareas_row_pct, pow_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, dtas_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, transition_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, permaculture_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, grocery_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, pubs_scenicareas_row_pct)
colnames(scenicareas_counts_pct) <- c("% Within scenicareass", "% within 50m", "% within 500m")
# Merge into larger dataframe
@ -1199,7 +1294,7 @@ Chasing down a curiosity, I decided to try and calculate whether proximity to "w
Proximity to these areas was the next concern, because many of these designations deliberately exclude human habitat, so it was necessary to measure the number of sites within proximity. There is a question which lies here regarding aesthetics, namely, what sort of proximity might generate an affective connection? From my own experience, I decided upon the distance represented by a short walk, i.e. a half-kilometre. However, with the more generic measurements, such as SSSI and forestation, this wouldn't do, as there are so many of these sites that a buffer of 500 meters encapsulates almost all of inhabited Scotland. So for these sites I also calculated a count within 50 metres.
So what did I discover? The results were inconclusive. First, it is important to note that on the whole, Eco-Congregations tend to be more urban than place of worship taken generally at a rate of nearly 3:1 (5.4% of Eco-Congregations lie in areas currently designated as "Very Remote Rural Areas" whereas nearly 15% of places of worship lie in these areas), so what I was testing for was whether this gap was smaller when specifying these various forms of "wild" remoteness. For our narrowest measurements, there were so few sites captured as to render measurement unreliable. There are, for obvious reasons, `st_within(ecs_sf, wildland)` sites located within any of SNG's core wild areas. Similarly, there are very few of our activist communities located within SSSI's (only `st_within(pow_pointX_sf, sssi)` places of worship out of `r length(pow_pointX)`, `st_within(transition_sf, sssi)` transition towns, (or 2%) and `st_within(dtas_sf, sssi)` community development trusts (3%)). However, expanding this out makes things a bit more interesting, within 50 metres of SSSI's in Scotland lie `st_within(ecs_sf, sssi_buf50)` Eco-Congregations (or just under 1%), which compares favourably with the `st_within(pow_pointX_sf, sssi_buf50)` places of worship (or just 1.5%) far exceeding our ratio (1:1.5 vs. 1:3). This is the same with our more anachronistic measure of "scenic areas," there are 7 eco-congregations within these areas, and 175 places of worship, making for a ratio of nearly 1:2 (2.1% vs. 4.3%). Taking our final measure, of forested areas, this is hard to calculate, as only `st_within(ecs_sf, forestinv)` Eco-Congregation lies within either native or generally forested land.
So what did I discover? The results were inconclusive. First, it is important to note that on the whole, Eco-Congregations tend to be more urban than place of worship taken generally at a rate of nearly 3:1 (5.4% of Eco-Congregations lie in areas currently designated as "Very Remote Rural Areas" whereas nearly 15% of places of worship lie in these areas), so what I was testing for was whether this gap was smaller when specifying these various forms of "wild" remoteness. For our narrowest measurements, there were so few sites captured as to render measurement unreliable. There are, for obvious reasons, `st_within(ecs, wildland)` sites located within any of SNG's core wild areas. Similarly, there are very few of our activist communities located within SSSI's (only `st_within(pow_pointX, sssi)` places of worship out of `r length(pow_pointX)`, `st_within(transition, sssi)` transition towns, (or 2%) and `st_within(dtas, sssi)` community development trusts (3%)). However, expanding this out makes things a bit more interesting, within 50 metres of SSSI's in Scotland lie `st_within(ecs, sssi_buf50)` Eco-Congregations (or just under 1%), which compares favourably with the `st_within(pow_pointX, sssi_buf50)` places of worship (or just 1.5%) far exceeding our ratio (1:1.5 vs. 1:3). This is the same with our more anachronistic measure of "scenic areas," there are 7 eco-congregations within these areas, and 175 places of worship, making for a ratio of nearly 1:2 (2.1% vs. 4.3%). Taking our final measure, of forested areas, this is hard to calculate, as only `st_within(ecs, forestinv)` Eco-Congregation lies within either native or generally forested land.
```{r 13_wilderness_tables}
@ -1241,11 +1336,11 @@ tm_shape(sssi_simplified, bbox = scotland) +
title = "Sites of Special Scientific Interest and ECS Groups") +
# tm_shape(sssi_buf50) + tm_borders(lwd=0.001) +
# tm_shape(sssi_buf500) + tm_borders(lwd=0.001)
tm_shape(admin_lev1_sf) + tm_borders(lwd=0.01) +
tm_shape(ecs_sf) + tm_dots("red", size = .02, alpha = .4) +
# tm_shape(ecs_sf_sssi50m) + tm_dots("yellow", size = .5, alpha = .4) +
# tm_shape(ecs_sf_sssi500m) + tm_dots("orange", size = .5, alpha = .4) +
# tm_shape(ecs_sf_sssibeyond500m) + tm_dots("red", size = .5, alpha = .4)
tm_shape(admin_lev1) + tm_borders(lwd=0.01) +
tm_shape(ecs) + tm_dots("red", size = .02, alpha = .4) +
# tm_shape(ecs_sssi50m) + tm_dots("yellow", size = .5, alpha = .4) +
# tm_shape(ecs_sssi500m) + tm_dots("orange", size = .5, alpha = .4) +
# tm_shape(ecs_sssibeyond500m) + tm_dots("red", size = .5, alpha = .4)
# tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
tm_credits("Data: UK Data Service (OGL)\n& Jeremy H. Kidwell,\nGraphic is CC-by-SA 4.0",
@ -1272,11 +1367,11 @@ tm_shape(sssi_simplified, bbox = scotland) + tm_fill(col = "blue", alpha = 0.4,
tm_shape(scenicareas_simplified, bbox = scotland) + tm_fill(col = "orange", alpha = 0.4, lwd=0.01) +
tm_shape(admin_lev1) + tm_borders(lwd=0.01) +
# tm_scale_bar(breaks = c(0, 100, 200), size = 1) +
tm_shape(ecs_sf) + tm_dots("red", size = .02, alpha = .4) +
tm_shape(pow_pointX_sf) + tm_dots("orange", size = .01, alpha = .2) +
tm_shape(dtas_sf) + tm_dots("blue", size = .02, alpha = .4) +
tm_shape(transition_sf) + tm_dots("green", size = .02, alpha = .4) +
tm_shape(permaculture_sf) + tm_dots("pink", size = .025, alpha = .4) +
tm_shape(ecs) + tm_dots("red", size = .02, alpha = .4) +
tm_shape(pow_pointX) + tm_dots("orange", size = .01, alpha = .2) +
tm_shape(dtas) + tm_dots("blue", size = .02, alpha = .4) +
tm_shape(transition) + tm_dots("green", size = .02, alpha = .4) +
tm_shape(permaculture) + tm_dots("pink", size = .025, alpha = .4) +
# tm_scale_bar(position = c("right", "bottom")) +
tm_style("gray") +
tm_credits("Data: UK Data Service (OGL)\n& Jeremy H. Kidwell,\nGraphic is CC-by-SA 4.0",
@ -1349,7 +1444,7 @@ urbanrural_table %>%
[^15541312]: This research was jointly funded by the AHRC/ESRC under project numnbers AH/K005456/1 and AH/P005063/1.
[^158261118]: This is not to say that there have been no collaborations before 2000, noteworthy in this respect is the WWF who helped to found the Alliance of Religion and Conservation (ARC) in 1985.
[^159141043]: This suggestion should be qualified - RSPB would greatly exceed ECS both in terms of the number of individual subscribers and budget. The RSPB trustee's report for 2013-2014 suggests that their member base was 1,114,938 people across Britain with a net income of £127m - the latter of which exceeds the Church of Scotland. If we adjust this based on the Scottish share of the population of the United Kingdom as of the 2011 census (8.3%) this leaves us with an income of £9.93m. The British charity commission requires charities to self-report the number of volunteers and staff, and from their most recent statistics we learn that RSPB engaged with 17,600 volunteers and employed 2,110 members of staff. Again, adjusted for population, this leaves 1,460 volunteers in Scotland and 176 staff. However, if we measure environmental groups based on the number of sites they maintain, RSPB has only 40 reserves with varying levels of local community engagement. For comparison, as of Sep 14 2015, Friends of the Earth Scotland had only 10 local groups (concentrated mostly in large urban areas). Depending on how one measures "volunteerism," it may be possible that ECS has more engaged volunteers in Scotland as well - if each ECS group had only 4 "volunteers" then this would exceed RSPB.
[^15541313]: Kidwell, Jeremy. (2016). Eco-Congregation Scotland, 2014-2016. University of Edinburgh. http://dx.doi.org/10.7488/ds/1357.
[^15541313]: Kidwell, Jeremy. (2016). Eco-Congregation Scotland, 2014-2016. University of Edinburgh. http://dx.doi.org/10.7488/ds/1357. Include citation here to RDJ article
[^15541342]:My dataset on transition towns will be made available later in 2016. Initial data was aquired from the Transition Scotland website http://www.transitionscotland.org/transition-in-scotland on December 10, 2014. We are currently in the process of collaboratively generating a more up-to-date dataset which will reflect their collaboration with SCCAN.
[^177171536]: For further detail on Dataset generation, see Kidwell, Forthcoming, 2018.
[^158261232]:Data was acquired from the Development Trusts Association website, http://www.dtascot.org.uk, accessed on 20 July 2015. As above, we are currently in the process of active collaboration with volunteers from the DTAS to co-generate a new dataset.
@ -1359,5 +1454,5 @@ urbanrural_table %>%
[^1554162]:From the Transition map key, "Green pins are 'official' groups
Blue pins are active communities who are connected to the Scottish Transition network Yellow pins show interest in this area"
[^15571030]:This was calculated by calculating a 10m wide footprint for every postcode in Scotland, areas which are not within 10m of a postcode (as of May 2014) are counted as uninhabited.
[^159142242]: Fiona Tweedie, *Ecumenical Audit: Questionnaire Findings* (2014).
[^159142242]: Fiona Tweedie, *Ecumenical Audit: Questionnaire Findings* (2014). [^footnote1]: Ref. IA article
[^15914204]:See note above regarding the data used from the PointX POI database. Note, for our research,we filtered out religious groups not represented within the Eco-Congregation footprint. We discuss representation by tradition and religion further below.adition and religion further below.

764
mapping_draft-hpc_optimised_wilderness.Rmd
View file

@ -1,764 +0,0 @@
---
title: "Mapping Environmental Action in Scotland"
abstract:
# thanks: "Replication files are available on the author's Github account (https://github.com/kidwellj/mapping_environmental_action). **Current version**: `r format(Sys.time(), '%B %d, %Y')`
style: jeremy1
author: "[Jeremy H. Kidwell](http://jeremykidwell.info)"
affiliation: University of Birmingham
institute: University of Birmingham
e-mail: "[j.kidwell@bham.ac.uk](mailto:j.kidwell@bham.ac.uk)"
date: "`r Sys.Date()`"
bibliography: biblio.bib
linkcolor: black
geometry: margin=1in
# fontfamily: mathpazo
fontsize: 11pt
output:
html_document:
theme: readable
keep_md: true
code_folding: hide
self_contained: true
toc: true
toc_depth: 2
number_sections: true
fig_caption: true
fig_retina: 1
pdf_document:
toc: false
keep_tex: true
number_sections: true
fig_caption: true
citation_package: natbib
latex_engine: xelatex
always_allow_html: yes
---
```{r setup, include=FALSE}
require(knitr)
require(kableExtra)
knitr::opts_chunk$set(fig.path='figures/', warning=FALSE, echo=FALSE, message=FALSE, dpi=300, fig.width=7)
# TODO: consider implementing knitcitations - https://github.com/cboettig/knitcitations
# TODO: fix simultaneous output towards PDF, see here: https://stackoverflow.com/questions/23621012/display-and-save-the-plot-simultaneously-in-r-rstudio
```
```{r load_packages, message=FALSE, warning=FALSE, include=FALSE}
## Default repo
# setwd("/Users/jeremy/gits/mapping_environmental_action")
# setwd("/Users/kidwellj/OneDrive\ -\ bham.ac.uk/writing/201708_mapping_environmental_action")
# Set repository to be new standard, e.g. cloud server.
# This will avoid a dialogue box if packages are to be installed for below on first run.
local({r <- getOption("repos")
r["CRAN"] <- "https://cloud.r-project.org"
options(repos=r)
})
# TODO: remove sp etc. once sf is fully implemented
# TODO: automatically test for packages below on given execution environment and run install.packages() as needed.
require(RCurl) # used for fetching reproducible datasets
require(sf) # new simplefeature data class, supercedes sp in many ways
require(sp) # needed for proj4string, deprecated by sf()
require(rgdal) # deprecated by sf()
require(GISTools) # deprecated by sf()
require(rgeos) # deprecated by sf()
require(maptools)
require(ggplot2)
require(tmap) # using as an alternative to base r graphics and ggplot for geospatial plots
require(tmaptools) # for get_asp_ratio below
require(grid) # using for inset maps on tmap
require(broom) # required for tidying SPDF to data.frame for ggplot2
require(tidyr) # using for grouped bar plot
require(plyr)
require(dplyr)
require(reshape2) # using for grouped bar plot
require(scales)
# require(sqldf) # using sqldf to filter before loading very large data sets
require(plotly) # allows for export of plots to dynamic web pages
require(gtable) # more powerful package for multi-plot layouts, not necessary for knitr
require(showtext) # for loading in fonts
require(extrafont) # font support
# Set up local workspace:
if (dir.exists("data") == FALSE) {
dir.create("data")
}
if (dir.exists("figures") == FALSE) {
dir.create("figures")
}
if (dir.exists("derivedData") == FALSE) {
dir.create("derivedData")
}
# Define Coordinate Reference Systems (CRS) for later use
# Note: I've used British National Grid (27000) in this paper, but have found that
# it is falling out of use in many cases, so will be defaulting to WGS84 in future
# data-sets and papers.
# Working with EPSG codes for spatialfeature CRS given the usage of this approach with sf()
bng <- CRS("+init=epsg:27700")
wgs84 <- CRS("+init=epsg:4326")
# Configure fonts for plots below
## Loading Google fonts (http://www.google.com/fonts)
# font_add_google("Merriweather", "merriweather")
# The following will load in system fonts (uncomment and run as needed on first execution)
# font_import(pattern="[A/a]rial", prompt=FALSE)
```
# Introduction[^15541312]
Until recently, environmentalism has been treated by governments and environmental charities as a largely secular concern. In spite of the well-developed tradition of "eco-theology" which began in earnest in the UK in the mid-twentieth century (and which has many precursors in previous centuries), third-sector groups and governments, particularly in Britain and Europe, have largely ignored religious groups as they have gone about their business crafting agendas for behaviour change, developing funding programmes, and developing platforms to mitigate ecological harm, motivate consumers and create regulation regimes. That this has changed is evidenced by the fact that several prominent non-religious environmental groups have commissioned studies and crafted outreach programmes to persons with a particular faith tradition or to "spiritual communities" including RSPB (2013) and the Sierra Club USA (2008).[^158261118] Further, since 2008, the Scottish Government has provided a significant portion of funding for the ecumenical charity, Eco-Congregation Scotland, which works to promote literacy on environmental issues in religious communities in Scotland and helps to certify congregations under their award programme. What is not well known, however, even by these religious environmental groups themselves, is whether or how their membership might be different from other environmental groups. This study represents an attempt to illuminate this new interest with some more concrete data about religious groups in Scotland and how they may differ from non-religious counterparts.
# Eco-Congregation Scotland: The Basics
```{r load_ecs_data, message=FALSE, warning=FALSE}
# read in Eco-Congregation Scotland data and-------------------
# ...turn it into a SpatialPointsDataFrame---------------------
# TODO: upload ECS-GIS-Locations_3.0.csv to zenodo repository, i.e.
ecs <- read.csv("data/ECS-GIS-Locations_3.0.csv", comment.char="#")
# unnecessary with advent of sf (above)
coordinates(ecs) <- c("X", "Y")
# Modified to use EPSG code directly 27 Feb 2019
proj4string(ecs) <- bng
ecs_sf <- st_as_sf(ecs, coords = c("X", "Y"), crs=paste0("+init=epsg:",27700))
```
There are `r length(ecs)` eco-congregations in Scotland. By some measurements, particularly in terms of individual sites and possibly also with regards to volunteers, this makes Eco-Congregation Scotland one of the largest environmental third-sector groups in Scotland.[^159141043]
In seeking to conduct GIS and statistical analysis of ECS, it is important to note that there some ways in which these sites are statistically opaque. Our research conducted through interviews at a sampling of sites and analysis of a variety of documents suggests that there is a high level of diversity both in terms of the number of those participating in environmental action and the types of action underway at specific sites. Work at a particular site can also ebb and flow over the course of time. Of course, as research into other forms of activism and secular environmental NGOs has shown, this is no different from any other third sector volunteer group. Variability is a regular feature of groups involved in activism and/or environmental concern.
For the sake of this analysis, we took each Eco-Congregation Scotland site to represent a point of analysis as if each specific site represented a community group which had "opted-in" on environmental concern. On this basis, in this section, in the tradition of human geography, we "map" environmental action among religious communities in Scotland a variety of ways. This is the first major geographical analysis of this kind conducted to date in Europe. We measure the frequency and location of ECS sites against a variety of standard geo-referenced statistical data sets, seeking to provide a statistical and geographically based assessment of the participation of religious groups in relation to the following:
- Location within Scotland
- Religious affiliation
- Relation to the Scottish Index of Multiple Deprivation (SIMD)
- Relation to the 8-Fold Scottish Government Urban-Rural Scale
- Proximity to "wilderness" (based on several different designations)
For the sake of comparison, we also measured the geographical footprint of two other forms of community group in Scotland, (1) Transition Towns (taking into account their recent merge with Scotland Communities Climate Action Network) and (2) member groups of the Development Trust Association Scotland ("DTAS"). These two groups provide a helpful basis for comparison as they are not centralised and thus have a significant geographical dispersion across Scotland. They also provide a useful comparison as transition is a (mostly) non-religious environmental movement, and community development trusts are not explicitly linked to environmental conservation (though this is often part of their remit), so we have a non-religious point of comparison in Transition and a non-environmental point of comparison with DTAS
# Technical Background
Analysis was conducted using QGIS 2.8 and R `r getRversion()`, and data-sets were generated in CSV format.[^15541313] To begin with, I assembled a data set consisting of x and y coordinates for each congregation in Scotland and collated this against a variety of other specific data. Coordinates were checked by matching UK postcodes of individual congregations against geo-referencing data in the Office for National Statistics postcode database. In certain instances a single "congregation" is actually a series of sites which have joined together under one administrative unit. In these cases, each site was treated as a separate data point if worship was held at that site at least once a month, but all joined sites shared a single unique identifier. As noted above, two other datasets were generated for the sake of comparative analysis.[^177171536] These included one similar Environmental Non-Governmental Organisation (ENGO) in Scotland (1) Transition Scotland (which includes Scotland Communities Climate Action Network);[^15541342] and another community-based NGO, Scottish Community Development Trusts.[^158261232] As this report will detail, these three overlap in certain instances both literally and in terms of their aims, but each also has a separate identity and footprint in Scotland. Finally, in order to normalise data, we utilised the PointX POI dataset which maintains a complete database of Places of Worship in Scotland.[^15541614]
# Background and History of Eco-Congregation Scotland
Eco-Congregation Scotland began a year before the official launch of Eco-Congregation England and Wales, in 1999, as part of an effort by Kippen Environment Centre (later renamed to Forth Environment Link, or "FEL") a charity devoted to environmental education in central Scotland[^158261210] to broaden the scope of its environmental outreach to churches in central Scotland.[^15826124] Initial funding was provided, through Kippen Environment Centre by way of a "sustainable action grant" (with funds drawn from a government landfill tax) through a government programme called Keep Scotland Beautiful (the Scottish cousin of Keep Britain Tidy). After this initial pilot project concluded, the Church of Scotland provided additional funding for the project in the form of staff time and office space. Additional funding a few years later from the Scottish Government helped subsidise the position of a business manager, and in 2011 the United Reformed Church contributed additional funding which subsidised the position of a full-time environmental chaplain for a 5-year term, bringing the total staff to five.
```{r calculate_ecs_by_year, message=FALSE, warning=FALSE}
# Tidy up date fields and convert to date data type
ecs$registration <- as.Date(ecs$registration, "%Y-%m-%d")
# TODO: Fix issues here with R complaining that "character string is not in a std...
# ecs$award1 <- as.Date(ecs$award1)
# ecs$award2 <- as.Date(ecs$award2)
# ecs$award3 <- as.Date(ecs$award3)
# ecs$award4 <- as.Date(ecs$award4)
# TODO: add "R" to command in paragraph below once this is resolved, do search for all non 'r instances
ecs_complete_cases <- ecs[complete.cases(ecs$year_begun),]
```
The programme launched officially in 2001 at Dunblane Cathedral in Stirling and by 2005 the project had `r length(ecs_complete_cases[ecs_complete_cases$year_begun < 2006, ])` congregations registered to be a part of the programme and 25 which had completed the curriculum successfully and received an Eco-Congregation award. By 2011, the number of registrations had tripled to `r length(ecs_complete_cases[ecs_complete_cases$year_begun < 2012, ])` and the number of awarded congregations had quadrupled to `sum(ecs$award1 < "01/01/2012", na.rm=TRUE)`. This process of taking registrations and using a tiered award or recognition scheme is common to many voluntary organisations. The ECS curriculum was developed in part by consulting the Eco-Congregation England and Wales materials which had been released just a year earlier in 1999, though it has been subsequently revised, particularly with a major redesign in 2010. In the USA, a number of similar groups take a similar approach including Earth Ministry (earthministry.org) and Green Faith (greenfaith.org).
In the case of Eco-Congregation Scotland, congregations are invited to begin by "registering" their interest in the programme by completing a basic one-sided form. The next step requires the completion of an award application, which includes a facilitated curriculum called a "church check-up" and after an application is submitted, the site is visited and assessed by third-party volunteer assessors. Sites are invited to complete additional applications for further awards which are incremental (as is the application process). Transition communities, at least in the period reflected on their map, go through a similar process (though this does not involve the use of a supplied curriculum) by which they are marked first as "interested," become "active" and then gain "official" status.[^1554162]
# Representation by Regional Authorities (Council Areas) {.tabset}
```{r import_admin_data, message=FALSE, warning=FALSE, include=FALSE}
# read in polygon for Scottish admin boundaries
# TODO: upload bundle of admin data to new zenodo repository and alter below to use new URLs
# TODO: need to remove readOGR below once st_read is confirmed to be working as sf
if (file.exists("data/scotland_ca_2010.shp") == FALSE) {
download.file("https://borders.ukdataservice.ac.uk/ukborders/easy_download/prebuilt/shape/Scotland_ca_2010.zip",
destfile = "data/Scotland_ca_2010.zip")
unzip("data/Scotland_ca_2010.zip", exdir = "data")
}
admin_lev1 <- readOGR("./data", "scotland_ca_2010")
admin_lev1_sf <- st_read("data/scotland_ca_2010.shp") %>% st_transform(paste0("+init=epsg:",27700))
# read in polygon for intermediate admin boundary layers
if (file.exists("data/scotland_parlcon_2011.shp") == FALSE) {
download.file("http://census.edina.ac.uk/ukborders/easy_download/prebuilt/shape/Scotland_parlcon_2011.zip",
destfile = "data/Scotland_parlcon_2011.zip")
unzip("data/Scotland_parlcon_2011.zip", exdir = "data")
}
admin_lev2 <- readOGR("./data", "scotland_parlcon_2011")
admin_lev2_sf <- st_read("data/scotland_parlcon_2011.shp") %>% st_transform(paste0("+init=epsg:",27700))
# Set CRS using epsg code on spdf for symmetry with datasets below
proj4string(admin_lev1) <- bng
proj4string(admin_lev2) <- bng
# Generate new sf shape using bounding box for central belt for map insets below
# Note: coordinates use BNG as CRS (EPSG: 27700)
scotland <- st_bbox(c(xmin = 5513.0000, xmax = 470332.0000,
ymin = 530249.0000, ymax = 1220301.5000),
crs = st_crs("+init=epsg:27700")) %>%
st_as_sfc()
centralbelt_region <- st_bbox(c(xmin = 234841, xmax = 346309,
ymin = 653542, ymax = 686722),
crs = st_crs("+init=epsg:27700")) %>%
st_as_sfc()
centralbelt_ratio <- get_asp_ratio(centralbelt_region)
scotland_ratio<- get_asp_ratio(scotland)
```
```{r import_groups_data, message=FALSE, warning=FALSE, include=FALSE}
# read in Transition Towns data and turn it into a SpatialPointsDataFrame
transition_wgs <- read.csv(text=getURL("https://zenodo.org/record/165519/files/SCCAN_1.4.csv"))
coordinates(transition_wgs) <- c("X", "Y")
proj4string(transition_wgs) <- CRS("+init=epsg:4326")
transition <- spTransform(transition_wgs, bng)
transition_sf <- st_as_sf(transition, coords = c("X", "Y"), crs=27700)
# read in all_churches data (data set generated by Jeremy Kidwell to replace PointX data used from Ordnance Survey)
# TODO: need to remove all data points which are outside BNG area to resolve error
# also need to make symmetrical with ECS denominations, add Methodist
# churches, remove nazarene and salvation army
# churches_all <- read.csv("data/all_churches_0.9.csv")
# churches_all_clean<-churches_all[complete.cases(churches_all),]
# churches_all_null<-churches_all[!complete.cases(churches_all),]
# coordinates(churches_all) <- c("X", "Y")
# proj4string(churches_all) <- proj4string(admin_lev1)
# read in pointX data and turn it into a SpatialPointsDataFrame
pow_pointX <- read.csv("./data/poi_2015_12_scot06340459.csv", sep="|")
coordinates(pow_pointX) <- c("feature_easting", "feature_northing")
# TODO: need to alter to draw from wgs84 or bng defined in preamble above
proj4string(pow_pointX) <- CRS("+init=epsg:27700")
pow_pointX_sf <- st_as_sf(pow_pointX, coords = c("X", "Y"), crs=27700)
# read in Scottish Community Dev. trust data and turn it into a SpatialPointsDataFrame
dtas <- read.csv("data/community-dev-trusts-2.6.csv")
coordinates(dtas) <- c("X", "Y")
proj4string(dtas) <- CRS("+init=epsg:27700")
dtas_sf <- st_as_sf(dtas, coords = c("X", "Y"), crs=27700)
# read in permaculture data and turn it into a SpatialPointsDataFrame
permaculture <- read.csv("data/permaculture_scot-0.8.csv")
coordinates(permaculture) <- c("X", "Y")
proj4string(permaculture) <- CRS("+init=epsg:27700")
permaculture_sf <- st_as_sf(permaculture, coords = c("X", "Y"), crs=27700)
# subset point datasets for inset maps below
ecs_sf_centralbelt <- st_intersection(ecs_sf, centralbelt_region)
dtas_sf_centralbelt <- st_intersection(dtas_sf, centralbelt_region)
transition_sf_centralbelt <- st_intersection(transition_sf, centralbelt_region)
pow_sf_centralbelt <- st_intersection(pow_pointX_sf, centralbelt_region)
permaculture_sf_centralbelt <- st_intersection(permaculture_sf, centralbelt_region)
```
```{r process_admin_data}
# Augment existing dataframes to run calculations and add columns with point counts per polygon,
# percentages, and normalising data.
# This code will generate a table of frequencies for each spatialpointsdataframe in admin
# calculate count of ECS for fields in admin and provide percentages
# JK Note: need to convert from poly.counts, which uses sp() to st_covers() which uses sf() - cf. https://stackoverflow.com/questions/45314094/equivalent-of-poly-counts-to-count-lat-long-pairs-falling-inside-of-polygons-w#45337050
admin_lev1$ecs_count <- poly.counts(ecs,admin_lev1)
admin_lev1$ecs_percent<- prop.table(admin_lev1$ecs_count)
# calculate count of places of worship in PointX db for fields in admin and provide percentages
admin_lev1$pow_count <- poly.counts(pow_pointX,admin_lev1)
admin_lev1$pow_percent<- prop.table(admin_lev1$pow_count)
# calculate count of Transition for fields in admin and provide percentages
admin_lev1$transition_count <- poly.counts(transition,admin_lev1)
admin_lev1$transition_percent<- prop.table(admin_lev1$transition_count)
# calculate count of dtas for fields in admin and provide percentages
admin_lev1$dtas_count <- poly.counts(dtas,admin_lev1)
admin_lev1$dtas_percent<- prop.table(admin_lev1$dtas_count)
# calculate count of permaculture for fields in admin and provide percentages
admin_lev1$permaculture_count <- poly.counts(permaculture,admin_lev1)
admin_lev1$permaculture_percent<- prop.table(admin_lev1$permaculture_count)
# run totals for intermediate boundaries level 2
# This code will generate a table of frequencies for each spatialpointsdataframe in admin_lev2
# calculate count of ECS for fields in admin_lev2 and provide percentages
admin_lev2$ecs_count <- poly.counts(ecs,admin_lev2)
admin_lev2$ecs_percent<- prop.table(admin_lev2$ecs_count)
# calculate count of places of worship in PointX db for fields in admin_lev2 and provide percentages
admin_lev2$pow_count <- poly.counts(pow_pointX,admin_lev2)
admin_lev2$pow_percent<- prop.table(admin_lev2$pow_count)
# calculate count of Transition for fields in admin_lev2 and provide percentages
admin_lev2$transition_count <- poly.counts(transition,admin_lev2)
admin_lev2$transition_percent<- prop.table(admin_lev2$transition_count)
# calculate count of dtas for fields in admin_lev2 and provide percentages
admin_lev2$dtas_count <- poly.counts(dtas,admin_lev2)
admin_lev2$dtas_percent<- prop.table(admin_lev2$dtas_count)
# calculate count of permaculture for fields in admin_lev2 and provide percentages
admin_lev2$permaculture_count <- poly.counts(permaculture,admin_lev2)
admin_lev2$permaculture_percent<- prop.table(admin_lev2$permaculture_count)
# calculate count of ECS for fields in admin_lev2_sf
# TODO: for future migration to sf throughout, remove above content and swap out references.
admin_lev2_sf$ecs_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(ecs_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$pow_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(pow_pointX_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$transition_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(transition_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$dtas_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(dtas_sf, coords = c("long", "lat"), crs = st_crs(27700))))
admin_lev2_sf$permaculture_count <- lengths(st_covers(admin_lev2_sf, st_as_sf(permaculture_sf, coords = c("long", "lat"), crs = st_crs(27700))))
# Import csv with population data for each level of administrative subdivision and join to spatialdataframe
# Placeholder for England parish data for later implementation
# download.file("http://census.edina.ac.uk/ukborders/easy_download/prebuilt/shape/England_cp_1991.zip", destfile = "parishes/parishes-1991.zip")
# unzip("parishes/parishes-1991.zip", exdir = "parishes")
# parishes <- rgdal::readOGR(dsn = "parishes", "england_cp_1991")
# TODO - consider adapting to use ONS mid-year statistics: https://www.ons.gov.uk/peoplepopulationandcommunity/populationandmigration/populationestimates/datasets/populationestimatesforukenglandandwalesscotlandandnorthernireland
# Load population statistics for normalising data by population and merge with admin_lev1
admin_lev1_pop <- read.csv("./data/scotland_admin_2011pop.csv")
admin_lev1 <- merge(x=admin_lev1, y=admin_lev1_pop, by.x = "code", by.y = "CODE")
# Convert number stored as string with a comma to an integer
admin_lev1$X2011_pop <- as.numeric(gsub(",", "", admin_lev1@data$X2011_pop))
# Calculate counts as percentages of total for normalising below
admin_lev1$pop_percent<- prop.table(admin_lev1$X2011_pop)
admin_lev1$pow_percent<- prop.table(admin_lev1$pow_count)
# Normalise ecs_count using population figures (using ArcGIS method)
admin_lev1$ecs_count_popnorm <- admin_lev1$ecs_count * admin_lev1$pop_percent
# Normalise ecs_count using places of worship counts (using ArcGIS method)
admin_lev1$ecs_count_pownorm <- admin_lev1$ecs_count * admin_lev1$pow_percent
# Preserve scale
admin_lev1$ecs_count_popnorm_scaled <- admin_lev1$ecs_count_popnorm*(sum(admin_lev1$ecs_count)/sum(admin_lev1$ecs_count_popnorm))
admin_lev1$ecs_count_pownorm_scaled <- admin_lev1$ecs_count_pownorm*(sum(admin_lev1$ecs_count)/sum(admin_lev1$ecs_count_pownorm))
# Load population statistics for normalising data by population on admin_lev2
# TODO: source this data file in a replicable way (e.g. zenodo)
admin_lev2_pop <- read.csv("./data/scotland_and_wales_const_scotland_2011pop.csv")
admin_lev2 <- merge(x=admin_lev2, y=admin_lev2_pop, by.x = "code", by.y = "CODE")
# Convert number stored as string with a comma to an integer (using data.zone data)
admin_lev2$Data.zone.Population <- as.numeric(gsub(",", "", admin_lev2$Data.zone.Population))
# Calculate counts as percentages of total for normalising below
admin_lev2$pop_percent <- prop.table(admin_lev2$Data.zone.Population)
admin_lev2$pow_percent <- prop.table(admin_lev2$pow_count)
# Normalise ecs_count using population figures (using ArcGIS method)
admin_lev2$ecs_count_popnorm <- admin_lev2$ecs_count * admin_lev2$pop_percent
# Normalise ecs_count using places of worship counts (using ArcGIS method)
admin_lev2$ecs_count_pownorm <- admin_lev2$ecs_count * admin_lev2$pow_percent
# Preserve scale
admin_lev2$ecs_count_popnorm_scaled <- admin_lev2$ecs_count_popnorm*(sum(admin_lev2$ecs_count)/sum(admin_lev2$ecs_count_popnorm))
admin_lev2$ecs_count_pownorm_scaled <- admin_lev2$ecs_count_pownorm*(sum(admin_lev2$ecs_count)/sum(admin_lev2$ecs_count_pownorm))
```
```{r additional_poi_data}
# Load in updated DTAS data set
dtas_new <- read.csv("data/dtas_4.0.csv")
coordinates(dtas_new) <- c("X", "Y")
proj4string(dtas_new) <- CRS("+init=epsg:27700")
# Load in retail data from geolytics dataset
# from here: https://geolytix.co.uk/?retail_points
poi_grocery_wgs <- read.csv("data/retailpoints_version11_dec17.txt", sep = "\t")
# select useful columns
poi_grocery_wgs <- subset(poi_grocery_wgs, select = c("retailer", "store_name", "long_wgs", "lat_wgs"))
# convert to spdf
coordinates(poi_grocery_wgs) <- c("long_wgs", "lat_wgs")
proj4string(poi_grocery_wgs) <- CRS("+init=epsg:4326")
poi_grocery <- spTransform(poi_grocery_wgs, proj4string(admin_lev1))
# filter out non-Scottish data
poi_grocery <- poi_grocery[!is.na(over(poi_grocery, geometry(admin_lev1))),]
poi_grocery_sf <- st_as_sf(poi_grocery, coords = c("long_wgs", "lat_wgs"), crs=paste0("+init=epsg:",27700))
# Load in British pubs from Ordnance survey dataset
poi_pubs <- read.csv("data/poi_pubs.csv", header = FALSE, sep = "|")
# select useful columns
poi_pubs <- subset(poi_pubs, select = c("V1", "V2", "V3", "V4", "V5"))
# rename columns to tidier names
colnames(poi_pubs) <- c("refnum", "name", "code", "x", "y")
coordinates(poi_pubs) <- c("x", "y")
proj4string(poi_pubs) <- proj4string(admin_lev1)
# filter out non-Scottish pubs
poi_pubs <- poi_pubs[!is.na(over(poi_pubs, geometry(admin_lev1))),]
poi_pubs_sf <- st_as_sf(poi_pubs, coords = c("x", "y"), crs=paste0("+init=epsg:",27700))
```
```{r wilderness_data_prep}
# 1. Download data for SSSI:
if (file.exists("data/SSSI_SCOTLAND.shp") == FALSE) {
# TODO: upload data to zenodo, uncomment below
# http://gateway.snh.gov.uk/natural-spaces/dataset.jsp?dsid=SSSI
# download.file("", destfile = "data/SSSI_SCOTLAND_ESRI.zip")
unzip("data/SSSI_SCOTLAND_ESRI.zip", exdir = "data")
}
sssi <- st_read("data/SSSI_SCOTLAND.shp") %>% st_transform(paste0("+init=epsg:",27700))
sssi_sp <- readOGR("./data", "SSSI_SCOTLAND")
# Generate simplified polygon for plots below
sssi_simplified <- st_simplify(sssi)
# sssi_simplified_sp <- rgeos::gSimplify(sssi_sp, tol=3)
# 2. Download wild land areas:
if (file.exists("data/WILDLAND_SCOTLAND.shp") == FALSE) {
# TODO: upload data to zenodo, uncomment below
# https://gateway.snh.gov.uk/natural-spaces/dataset.jsp?dsid=WILDLAND
# download.file("", destfile = "data/WILDLAND_SCOTLAND_ESRI.zip")
unzip("data/WILDLAND_SCOTLAND_ESRI.zip", exdir = "data")
}
wildland <- st_read("data/WILDLAND_SCOTLAND.shp") %>% st_transform(paste0("+init=epsg:",27700))
# Generate simplified polygon for plots below
wildland_simplified <- st_simplify(wildland)
# 3. Download data for National Forest Inventory:
# Note: UK-wide data is here: https://opendata.arcgis.com/datasets/bcd6742a2add4b68962aec073ab44138_0.zip?outSR=%7B%22wkid%22%3A27700%2C%22latestWkid%22%3A27700%7D
if (file.exists("data/National_Forest_Inventory_Woodland_Scotland_2017.shp") == FALSE) {
download.file("https://opendata.arcgis.com/datasets/3cb1abc185a247a48b9d53e4c4a8be87_0.zip?outSR=%7B%22wkid%22%3A27700%2C%22latestWkid%22%3A27700%7D",
destfile = "data/National_Forest_Inventory_Woodland_Scotland_2017.zip")
unzip("data/National_Forest_Inventory_Woodland_Scotland_2017.zip", exdir = "data")
}
forestinv <- st_read("data/National_Forest_Inventory_Woodland_Scotland_2017.shp") %>% st_transform(paste0("+init=epsg:",27700))
# Generate simplified polygon for plots below
forestinv_simplified <- st_simplify(forestinv)
# Download data for scenic areas
# https://opendata.arcgis.com/datasets/d7f6b987c7224a72a185ce012258d500_23.zip
# International Union for Conservation of Nature (IUCN), Category V Protected Landscapes
# for England: https://environment.data.gov.uk/DefraDataDownload/?mapService=NE/AreasOfOutstandingNaturalBeautyEngland&Mode=spatial
# download.file("https://opendata.arcgis.com/datasets/d7f6b987c7224a72a185ce012258d500_23.zip", destfile = "data/ScenicAreas.zip")
# unzip("data/ScenicAreas.zip", exdir = "data")
scenicareas <- st_read("data/SG_NationalScenicAreas_1998.shp") %>% st_transform(paste0("+init=epsg:",27700))
# Generate simplified polygon for plots below
scenicareas_simplified <- st_simplify(scenicareas)
# Set symmetrical CRS for analysis below (inserted here in order to correct errors, may be deprecated later)
# st_crs(sssi) <- 27700
# st_crs(ecs_sf) <- 27700
# st_crs(pow_pointX_sf) <- 27700
# st_crs(dtas_sf) <- 27700
# st_crs(transition_sf) <- 27700
# st_crs(permaculture_sf) <- 27700
# Define buffer and measure number of ECS groups within 0.5 miles of all SSSI
# CRS uses meters for units, so a buffer for 0.5 miles would use 805 meters)
# Define buffers as lines and polygons for reuse below (as some of these operations take > 10 minutes)
# TODO: Working with simplified polygons here to optimise execution, need to confirm calculations are the same as polygons without simplification (likely!)
sssi_buf50 <- st_buffer(sssi_simplified, dist = 50)
sssi_buf500 <- st_buffer(sssi_simplified, dist = 500)
# Lines offer even more optimised representation suitable for plotting
# TODO: resolve error:
# Error in st_cast.sfc(., to = "LINESTRING") :
# use smaller steps for st_cast; first cast to MULTILINESTRING or POLYGON?
# sssi_buf50_lines <- st_union(sssi_simplified) %>% st_buffer(50) %>%
# st_cast(to = "MULTILINESTRING") %>%
# st_cast(to = "LINESTRING")
# sssi_buf500_lines <- st_union(sssi_simplified) %>% st_buffer(500) %>%
# st_cast(to = "MULTILINESTRING") %>%
# st_cast(to = "LINESTRING")
wildland_buf50 <- st_buffer(wildland_simplified, dist = 50)
wildland_buf500 <- st_buffer(wildland_simplified, dist = 500)
# Lines offer even more optimised representation suitable for plotting
# wildland_buf50_lines = st_union(wildland_simplified) %>% st_buffer(50) %>%
# st_cast(to = "LINESTRING")
# wildland_buf500_lines = st_union(wildland_simplified) %>% st_buffer(500) %>%
# st_cast(to = "LINESTRING")
forestinv_buf50 <- st_buffer(forestinv_simplified, dist = 50)
forestinv_buf500 <- st_buffer(forestinv_simplified, dist = 500)
# Lines offer even more optimised representation suitable for plotting
# forestinv_buf50_lines = st_union(forestinv_simplified) %>% st_buffer(50) %>%
# st_cast(to = "LINESTRING")
# forestinv_buf500_lines = st_union(forestinv_simplified) %>% st_buffer(500) %>%
# st_cast(to = "LINESTRING")
scenicareas_buf50 <- st_buffer(scenicareas_simplified, dist = 50)
scenicareas_buf500 <- st_buffer(scenicareas_simplified, dist = 500)
# Calculate number of groups within polygons
# calculate coincidence of ecs points within each polygons and buffers for each
# TODO: possibly use st_difference(sssi_buf50, sssi)
# TODO: consider subsetting instead here, e.g.:
# plot(lnd, col = "lightgrey") # plot the london_sport object
# sel <- lnd$Partic_Per > 25
# plot(lnd[ sel, ], col = "turquoise", add = TRUE) # add selected zones to map
# from https://gotellilab.github.io/Bio381/StudentPresentations/SpatialDataTutorial.html
# TODO: integrate pre-calc here into calculations further down which are still recalculating these figures
ecs_sf_sssi <- st_within(ecs_sf, sssi_simplified)
ecs_sf_sssi50m <- st_within(ecs_sf, sssi_buf50)
ecs_sf_sssi500m <- st_within(ecs_sf, sssi_buf500)
ecs_sf_sssibeyond500m <- !(st_within(ecs_sf, sssi_buf500))
ecs_sf_wildland <- st_within(ecs_sf, wildland_simplified)
ecs_sf_wildland50m <- st_within(ecs_sf, wildland_buf50)
ecs_sf_wildland500m <- st_within(ecs_sf, wildland_buf500)
ecs_sf_wildlandbeyond500m <- !(st_within(ecs_sf, wildland_buf500))
ecs_sf_forestinv <- st_within(ecs_sf, forestinv_simplified)
ecs_sf_forestinv50m <- st_within(ecs_sf, forestinv_buf50)
ecs_sf_forestinv500m <- st_within(ecs_sf, forestinv_buf500)
ecs_sf_forestinvbeyond500m <- !(st_within(ecs_sf, forestinv_buf500))
ecs_sf_scenicareas <- st_within(ecs_sf, scenicareas_simplified)
ecs_sf_scenicareas50m <- st_within(ecs_sf, scenicareas_buf50)
ecs_sf_scenicareas500m <- st_within(ecs_sf, scenicareas_buf500)
ecs_sf_scenicareasbeyond500m <- !(st_within(ecs_sf, scenicareas_buf500))
# TODO: implement more efficient code using do.call() function or sapply() as here https://stackoverflow.com/questions/3642535/creating-an-r-dataframe-row-by-row
# TODO: implement parallel computing to distribute execution of loopable calculations below
# See: https://nceas.github.io/oss-lessons/parallel-computing-in-r/parallel-computing-in-r.html
# Generate dataframe based on SSSI buffers
# Calculate incidence of ecs within SSSI and within buffers at 50/500m
ecs_sssi_row <- c(sum(apply(st_within(ecs_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, sssi_buf500, sparse=FALSE), 1, any)))
pow_sssi_row <- c(sum(apply(st_within(pow_pointX_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, sssi_buf500, sparse=FALSE), 1, any)))
dtas_sssi_row <- c(sum(apply(st_within(dtas_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, sssi_buf500, sparse=FALSE), 1, any)))
transition_sssi_row <- c(sum(apply(st_within(transition_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, sssi_buf500, sparse=FALSE), 1, any)))
permaculture_sssi_row <- c(sum(apply(st_within(permaculture_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, sssi_buf500, sparse=FALSE), 1, any)))
grocery_sssi_row <- c(sum(apply(st_within(poi_grocery_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, sssi_buf500, sparse=FALSE), 1, any)))
pubs_sssi_row <- c(sum(apply(st_within(poi_pubs_sf, sssi_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, sssi_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, sssi_buf500, sparse=FALSE), 1, any)))
# Generate dataframe from rows based on counts
sssi_counts <- rbind(ecs_sssi_row, pow_sssi_row)
sssi_counts <- rbind(sssi_counts, dtas_sssi_row)
sssi_counts <- rbind(sssi_counts, transition_sssi_row)
sssi_counts <- rbind(sssi_counts, permaculture_sssi_row)
sssi_counts <- rbind(sssi_counts, grocery_sssi_row)
sssi_counts <- rbind(sssi_counts, pubs_sssi_row)
sssi_counts <- as.data.frame(sssi_counts)
colnames(sssi_counts) <- c("Within SSSIs", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
# TODO: Work with ecs_sf etc., but first sort out why length is different between shapefiles
ecs_sssi_row_pct <- ecs_sssi_row/length(ecs)
pow_sssi_row_pct <- pow_sssi_row/length(pow_pointX)
dtas_sssi_row_pct <- dtas_sssi_row/length(dtas)
transition_sssi_row_pct <- transition_sssi_row/length(transition)
permaculture_sssi_row_pct <- permaculture_sssi_row/length(permaculture)
grocery_sssi_row_pct <- grocery_sssi_row/length(poi_grocery)
pubs_sssi_row_pct <- pubs_sssi_row/length(poi_pubs)
sssi_counts_pct <- rbind(ecs_sssi_row_pct, pow_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, dtas_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, transition_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, permaculture_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, grocery_sssi_row_pct)
sssi_counts_pct <- rbind(sssi_counts_pct, pubs_sssi_row_pct)
colnames(sssi_counts_pct) <- c("% Within SSSIs", "% within 50m", "% within 500m")
# Merge into larger dataframe
sssi_counts_merged <- cbind(sssi_counts, sssi_counts_pct)
# Generate dataframe based on wildland buffers
ecs_wildland_row <- c(sum(apply(st_within(ecs_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, wildland_buf500, sparse=FALSE), 1, any)))
pow_wildland_row <- c(sum(apply(st_within(pow_pointX_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(ecs_wildland_row, pow_wildland_row)
dtas_wildland_row <- c(sum(apply(st_within(dtas_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, dtas_wildland_row)
transition_wildland_row <- c(sum(apply(st_within(transition_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, transition_wildland_row)
permaculture_wildland_row <- c(sum(apply(st_within(permaculture_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, permaculture_wildland_row)
grocery_wildland_row <- c(sum(apply(st_within(poi_grocery_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, grocery_wildland_row)
pubs_wildland_row <- c(sum(apply(st_within(poi_pubs_sf, wildland_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, wildland_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, wildland_buf500, sparse=FALSE), 1, any)))
wildland_counts <- rbind(wildland_counts, pubs_wildland_row)
colnames(wildland_counts) <- c("Within Wildland Areas", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
ecs_wildland_row_pct <- ecs_wildland_row/length(ecs)
pow_wildland_row_pct <- pow_wildland_row/length(pow_pointX)
dtas_wildland_row_pct <- dtas_wildland_row/length(dtas)
transition_wildland_row_pct <- transition_wildland_row/length(transition)
permaculture_wildland_row_pct <- permaculture_wildland_row/length(permaculture)
grocery_wildland_row_pct <- grocery_wildland_row/length(poi_grocery)
pubs_wildland_row_pct <- pubs_wildland_row/length(poi_pubs)
wildland_counts_pct <- rbind(ecs_wildland_row_pct, pow_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, dtas_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, transition_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, permaculture_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, grocery_wildland_row_pct)
wildland_counts_pct <- rbind(wildland_counts_pct, pubs_wildland_row_pct)
colnames(wildland_counts_pct) <- c("% Within wildlands", "% within 50m", "% within 500m")
# Merge into larger dataframe
wildland_counts_merged <- cbind(wildland_counts, wildland_counts_pct)
# Generate dataframe based on forestinv buffers
ecs_forestinv_row <- c(sum(apply(st_within(ecs_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, forestinv_buf500, sparse=FALSE), 1, any)))
pow_forestinv_row <- c(sum(apply(st_within(pow_pointX_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(ecs_forestinv_row, pow_forestinv_row)
dtas_forestinv_row <- c(sum(apply(st_within(dtas_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, dtas_forestinv_row)
transition_forestinv_row <- c(sum(apply(st_within(transition_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, transition_forestinv_row)
permaculture_forestinv_row <- c(sum(apply(st_within(permaculture_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, permaculture_forestinv_row)
grocery_forestinv_row <- c(sum(apply(st_within(poi_grocery_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, grocery_forestinv_row)
pubs_forestinv_row <- c(sum(apply(st_within(poi_pubs_sf, forestinv_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, forestinv_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, forestinv_buf500, sparse=FALSE), 1, any)))
forestinv_counts <- rbind(forestinv_counts, pubs_forestinv_row)
colnames(forestinv_counts) <- c("Within Woodlands", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
# TODO: fix error generated by ecs_forestinv_row_pct using ecs_sf. Migrate all these to sf, but check for errors.
ecs_forestinv_row_pct <- ecs_forestinv_row/length(ecs)
pow_forestinv_row_pct <- pow_forestinv_row/length(pow_pointX)
dtas_forestinv_row_pct <- dtas_forestinv_row/length(dtas)
transition_forestinv_row_pct <- transition_forestinv_row/length(transition)
permaculture_forestinv_row_pct <- permaculture_forestinv_row/length(permaculture)
grocery_forestinv_row_pct <- grocery_forestinv_row/length(poi_grocery)
pubs_forestinv_row_pct <- pubs_forestinv_row/length(poi_pubs)
forestinv_counts_pct <- rbind(ecs_forestinv_row_pct, pow_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, dtas_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, transition_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, permaculture_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, grocery_forestinv_row_pct)
forestinv_counts_pct <- rbind(forestinv_counts_pct, pubs_forestinv_row_pct)
colnames(forestinv_counts_pct) <- c("% Within Woodlands", "% within 50m", "% within 500m")
# Merge into larger dataframe
forestinv_counts_merged <- cbind(forestinv_counts, forestinv_counts_pct)
# Generate dataframe based on scenicareas buffers
ecs_scenicareas_row <- c(sum(apply(st_within(ecs_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(ecs_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
pow_scenicareas_row <- c(sum(apply(st_within(pow_pointX_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(pow_pointX_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(ecs_scenicareas_row, pow_scenicareas_row)
dtas_scenicareas_row <- c(sum(apply(st_within(dtas_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(dtas_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, dtas_scenicareas_row)
transition_scenicareas_row <- c(sum(apply(st_within(transition_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(transition_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, transition_scenicareas_row)
permaculture_scenicareas_row <- c(sum(apply(st_within(permaculture_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(permaculture_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, permaculture_scenicareas_row)
grocery_scenicareas_row <- c(sum(apply(st_within(poi_grocery_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_grocery_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, grocery_scenicareas_row)
pubs_scenicareas_row <- c(sum(apply(st_within(poi_pubs_sf, scenicareas_simplified, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, scenicareas_buf50, sparse=FALSE), 1, any)), sum(apply(st_within(poi_pubs_sf, scenicareas_buf500, sparse=FALSE), 1, any)))
scenicareas_counts <- rbind(scenicareas_counts, pubs_scenicareas_row)
colnames(scenicareas_counts) <- c("Within Scenic Areas", "...50m", "...500m")
# Generate dataframe from rows based on percentages of totals
ecs_scenicareas_row_pct <- ecs_scenicareas_row/length(ecs)
pow_scenicareas_row_pct <- pow_scenicareas_row/length(pow_pointX)
dtas_scenicareas_row_pct <- dtas_scenicareas_row/length(dtas)
transition_scenicareas_row_pct <- transition_scenicareas_row/length(transition)
permaculture_scenicareas_row_pct <- permaculture_scenicareas_row/length(permaculture)
grocery_scenicareas_row_pct <- grocery_scenicareas_row/length(poi_grocery)
pubs_scenicareas_row_pct <- pubs_scenicareas_row/length(poi_pubs)
scenicareas_counts_pct <- rbind(ecs_scenicareas_row_pct, pow_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, dtas_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, transition_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, permaculture_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, grocery_scenicareas_row_pct)
scenicareas_counts_pct <- rbind(scenicareas_counts_pct, pubs_scenicareas_row_pct)
colnames(scenicareas_counts_pct) <- c("% Within scenicareass", "% within 50m", "% within 500m")
# Merge into larger dataframe
scenicareas_counts_merged <- cbind(scenicareas_counts, scenicareas_counts_pct)
```
# Proximity to "Wilderness"
Chasing down a curiosity, I decided to try and calculate whether proximity to "wilderness" or "scenic nature" or just trees might have some impact on generating more mobilised communities. I realised that there would be several problems with this kind of calculation up front, first being that "nature" is a deeply problematic construct, reviled by geographers and philosophers alike. With this in mind, I identified several different ways of reckoning wilderness, starting with the highly anachronistic "Scenic Land" designation from the 1970s. Then I pursued the more carefully calculated "core wild areas" generated by SNH just a few years ago. However, even the core wile areas concept has been criticised heavily, so I also expanded out my search to include all sites of special scientific interest and then went even wider to include the Scottish Forestry Service's "Native Woodland" and finally, the most generic possible measurement, any land identified as forested at the last Forest Inventory.
Proximity to these areas was the next concern, because many of these designations deliberately exclude human habitat, so it was necessary to measure the number of sites within proximity. There is a question which lies here regarding aesthetics, namely, what sort of proximity might generate an affective connection? From my own experience, I decided upon the distance represented by a short walk, i.e. a half-kilometre. However, with the more generic measurements, such as SSSI and forestation, this wouldn't do, as there are so many of these sites that a buffer of 500 meters encapsulates almost all of inhabited Scotland. So for these sites I also calculated a count within 50 metres.
So what did I discover? The results were inconclusive. First, it is important to note that on the whole, Eco-Congregations tend to be more urban than place of worship taken generally at a rate of nearly 3:1 (5.4% of Eco-Congregations lie in areas currently designated as "Very Remote Rural Areas" whereas nearly 15% of places of worship lie in these areas), so what I was testing for was whether this gap was smaller when specifying these various forms of "wild" remoteness. For our narrowest measurements, there were so few sites captured as to render measurement unreliable. There are, for obvious reasons, `st_within(ecs_sf, wildland)` sites located within any of SNG's core wild areas. Similarly, there are very few of our activist communities located within SSSI's (only `st_within(pow_pointX_sf, sssi)` places of worship out of `r length(pow_pointX)`, `st_within(transition_sf, sssi)` transition towns, (or 2%) and `st_within(dtas_sf, sssi)` community development trusts (3%)). However, expanding this out makes things a bit more interesting, within 50 metres of SSSI's in Scotland lie `st_within(ecs_sf, sssi_buf50)` Eco-Congregations (or just under 1%), which compares favourably with the `st_within(pow_pointX_sf, sssi_buf50)` places of worship (or just 1.5%) far exceeding our ratio (1:1.5 vs. 1:3). This is the same with our more anachronistic measure of "scenic areas," there are 7 eco-congregations within these areas, and 175 places of worship, making for a ratio of nearly 1:2 (2.1% vs. 4.3%). Taking our final measure, of forested areas, this is hard to calculate, as only `st_within(ecs_sf, forestinv)` Eco-Congregation lies within either native or generally forested land.
```{r 13_wilderness_tables}
# Output mmd tables using kable
sssi_counts_merged %>%
kable(format = "html", col.names = colnames(sssi_counts_merged), caption = "Group counts within SSSIs") %>%
kable_styling(bootstrap_options = c("striped", "hover", "condensed", full_width = F, "responsive"))
wildland_counts_merged %>%
kable(format = "html", col.names = colnames(wildland_counts_merged), caption = "Group counts within Wildland Areas") %>%
kable_styling(bootstrap_options = c("striped", "hover", "condensed", full_width = F, "responsive"))
forestinv_counts_merged %>%
kable(format = "html", col.names = colnames(forestinv_counts_merged), caption = "Group counts within Woodlands") %>%
kable_styling(bootstrap_options = c("striped", "hover", "condensed", full_width = F, "responsive"))
scenicareas_counts_merged %>%
kable(format = "html", col.names = colnames(scenicareas_counts_merged), caption = "Group counts within Scenic Areas") %>%
kable_styling(bootstrap_options = c("striped", "hover", "condensed", full_width = F, "responsive"))
# Output CSV files for tables above
write.csv(sssi_counts_merged, "derivedData/sssi_counts_merged.csv", row.names=TRUE)
write.csv(wildland_counts_merged, "derivedData/wildland_counts_merged.csv", row.names=TRUE)
write.csv(forestinv_counts_merged, "derivedData/forestinv_counts_merged.csv", row.names=TRUE)
write.csv(scenicareas_counts_merged, "derivedData/scenicareas_counts_merged.csv", row.names=TRUE)
```
# Citations
[^15541312]: This research was jointly funded by the AHRC/ESRC under project numnbers AH/K005456/1 and AH/P005063/1.
[^158261118]: This is not to say that there have been no collaborations before 2000, noteworthy in this respect is the WWF who helped to found the Alliance of Religion and Conservation (ARC) in 1985.
[^159141043]: This suggestion should be qualified - RSPB would greatly exceed ECS both in terms of the number of individual subscribers and budget. The RSPB trustee's report for 2013-2014 suggests that their member base was 1,114,938 people across Britain with a net income of £127m - the latter of which exceeds the Church of Scotland. If we adjust this based on the Scottish share of the population of the United Kingdom as of the 2011 census (8.3%) this leaves us with an income of £9.93m. The British charity commission requires charities to self-report the number of volunteers and staff, and from their most recent statistics we learn that RSPB engaged with 17,600 volunteers and employed 2,110 members of staff. Again, adjusted for population, this leaves 1,460 volunteers in Scotland and 176 staff. However, if we measure environmental groups based on the number of sites they maintain, RSPB has only 40 reserves with varying levels of local community engagement. For comparison, as of Sep 14 2015, Friends of the Earth Scotland had only 10 local groups (concentrated mostly in large urban areas). Depending on how one measures "volunteerism," it may be possible that ECS has more engaged volunteers in Scotland as well - if each ECS group had only 4 "volunteers" then this would exceed RSPB.
[^15541313]: Kidwell, Jeremy. (2016). Eco-Congregation Scotland, 2014-2016. University of Edinburgh. http://dx.doi.org/10.7488/ds/1357.
[^15541342]:My dataset on transition towns will be made available later in 2016. Initial data was aquired from the Transition Scotland website http://www.transitionscotland.org/transition-in-scotland on December 10, 2014. We are currently in the process of collaboratively generating a more up-to-date dataset which will reflect their collaboration with SCCAN.
[^177171536]: For further detail on Dataset generation, see Kidwell, Forthcoming, 2018.
[^158261232]:Data was acquired from the Development Trusts Association website, http://www.dtascot.org.uk, accessed on 20 July 2015. As above, we are currently in the process of active collaboration with volunteers from the DTAS to co-generate a new dataset.
[^15541614]:PointX data for "Landscape Data" items is sourced from Ordnance Survey Land-Line and MasterMap(R) and the data points are augmented with additional information provided through research by PointX staff, and data aquired from unidentified "local data companie(s)" and the "118 Information" database (see: http://www.118information.co.uk). This data is under license and cannot be made available for use. It is important to note that I became aware of inaccuracies in this dataset over the course of use and subsequently generated my own dataset in collaboration with churches in Scotland. This will be made available later in 2016. I am in active conversation with OS about improving the quality of the data in PointX regarding places of worship.
[^15826124]:Interview with Margaret Warnock, 29 Aug 2014.
[^158261210]:From http://www.forthenvironmentlink.org, accessed 12 July 2015.
[^1554162]:From the Transition map key, "Green pins are 'official' groups
Blue pins are active communities who are connected to the Scottish Transition network Yellow pins show interest in this area"
[^15571030]:This was calculated by calculating a 10m wide footprint for every postcode in Scotland, areas which are not within 10m of a postcode (as of May 2014) are counted as uninhabited.
[^159142242]: Fiona Tweedie, *Ecumenical Audit: Questionnaire Findings* (2014).
[^15914204]:See note above regarding the data used from the PointX POI database. Note, for our research,we filtered out religious groups not represented within the Eco-Congregation footprint. We discuss representation by tradition and religion further below.adition and religion further below.