-
Notifications
You must be signed in to change notification settings - Fork 2
/
README.Rmd
220 lines (142 loc) · 11.8 KB
/
README.Rmd
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
---
output:
github_document:
pandoc_args: --webtex=https://latex.codecogs.com/svg.latex?
always_allow_html: true
---
```{r, include = FALSE}
knitr::opts_chunk$set(
collapse = TRUE,
comment = "#>",
fig.path = "man/figures/README-",
out.width = "100%"
)
```
# GVI: Greenness Visibility Index [<img src="man/figures/logo.png" align="right" height="139"/>](https://github.com/STBrinkmann/GVI)
[](https://doi.org/10.5281/zenodo.7057132) [](https://doi.org/10.5281/zenodo.5061257) [](https://github.com/jonnyhuck/green-visibility-index)
The `GVI` R package helps researchers compute the Greenness Visibility Index (GVI) presented by [Labib, Huck and Lindley (2021)](https://doi.org/10.1016/j.scitotenv.2020.143050) and includes optimisation of the underlying algortihm proposed by [Brinkmann et al. (2022)](https://doi.org/10.5194/agile-giss-3-27-2022). The GVI is calculated using a Digital Surface Model (DSM), Digital Terrain Model (DTM) and Greenness Raster. `GVI` is written in C++ to provide fast and light weighted functionality.
- [Functions](#functions)
- [Installation](#installation)
- [Methods](#methods)
- [Examples](#examples)
- [Citation](#citation)
GVI can also be computed in Python, by [Jonny Huck](https://github.com/jonnyhuck/green-visibility-index)
## Functions {#functions}
| Function | Description |
|-------------------------|-----------------------------------------------|
| [*viewshed*](#viewshed) | Computes the viewshed of a single point on a Digital Surface Model |
| [*visualizeWeights*](#visualize-weights) | Helper function, to adjust spatial weight parameters in the *vgvi* and *vgvi_from_sf* functions |
| [*vgvi_from_sf*](#viewshed-greenness-visibility-index-vgvi-from-sf) | Computes the *viewshed* and calculates the proportion of visible greenspace; Supports multiple points, lines or polygons |
| sf_to_rast | Function for computing a continuous raster map from a sf object by interpolating the sf values using Inverse Distance Weighting (IDW). |
## Installation {#installation}
1. Install [R](https://cran.r-project.org/)
- (Recommended) Install [RStudio](https://www.rstudio.com/products/rstudio/download/)
2. Install [terra](https://github.com/rspatial/terra)
3. Install GVI:
- Install remotes
`install.packages("remotes")`
- Install `GVI`
`GVI` is still in very active development. Therefore, the package is also not on CRAN yet. You can install the latest version of `GVI` from GitHub with:\
`remotes::install_git("https://github.com/STBrinkmann/GVI")`
## Methods {#methods}
### Viewshed {#viewshed}
The `viewshed` function computes a binary viewshed of a single point on a Digital Surface Model (DSM) raster. A radial buffer is applied on the observer position, and visibility is beeing calculated usig a C++ implementation of Bresenham's line algorithm [[Bresenham 1965](https://doi.org/10.1147/sj.41.0025), [Bresenham 1977](https://doi.org/10.1145/359423.359432)] and simple geometry. The result of the `viewshed` function is a radial raster where 0 = no-visible and 1 = visible area.
### Viewshed Greenness Visibility Index (VGVI)
The VGVI expresses the proportion of visible greenness to the total visible area based on a viewshed. The estimated VGVI values range between 0 and 1, where 0 = no green cells are visible, and 1 = all of the visible cells are green.
Based on a viewshed and a binary greenspace raster, all visible points are classified as visible green and visible no-green. All values are summarized using a decay function, to account for the reducing visual prominence of an object in space with increasing distance from the observer. Currently two options are supported, a logistic and an exponential function.
$$\begin{align*} f(x) = \cfrac{1}{1 + e^{b(x-m)}} && \text{(logistic)}\\ f(x) = \cfrac{1}{1 + (bx^{m})} && \text{(exponential)} \end{align*} $$
The full algorithm has been described in [Brinkmann et al. (2022)](https://doi.org/10.5194/agile-giss-3-27-2022).
### Visualize Weights {#visualize-weights}
The `visualizeWeights` function helps to adjust spatial weight parameters *m* and *b* used in the `vgvi` and `vgvi_from_sf` functions.
*Example output of the* `visualizeWeight` *function to compare and parameterize the decay weights of a logistic (left) and an exponential (right) function.*
### Viewshed Greenness Visibility Index (VGVI) from sf {#viewshed-greenness-visibility-index-vgvi-from-sf}
The `vgvi_from_sf` function combines the `viewshed` function to compute a raster of visible cells, and calculates the VGVI using a binary GreenSpace map. The input of the observer location is an [sf](https://r-spatial.github.io/sf/) object (supports points, lines and polygons). If this object is a line or polygon feature, it will be converted to points. For every point, first the viewshed, than the VGVI is being calculated. The output of the `vgvi_from_sf` function is a single sf point object containing the VGVI for every point. This function supports parallel computation on Windows, Linux and MacOS, however it is highly recommended to use Linux or MacOS.
## Examples {#examples}
### Data
For the first two examples we will use a [Digital Elevation Model (DEM)](https://opendata.vancouver.ca/explore/dataset/digital-elevation-model), a binary Greenspace Mask based on a [land cover classification](http://www.metrovancouver.org/data) and a Digital Surface Model (DSM). The DSM is generated from [LiDAR data collected in 2013](https://opendata.vancouver.ca/explore/dataset/lidar-2013/). A detailed explanation on how the DSM has been generated using R is provided in [this](https://geobrinkmann.com/post/visible-greenness-exposure/) tutorial. To reduce the size of the R package, the sample data has been uploaded to a separate GitHub repository and needs to be downloaded first.
```{r message=FALSE, warning=FALSE}
# Download DEM
DEM_tmp <- tempfile(fileext = ".tif")
download.file(url = "https://github.com/STBrinkmann/data/raw/main/GVI_Data/GVI_DEM.tif",
destfile = DEM_tmp, mode="wb")
# Download DSM
DSM_tmp <- tempfile(fileext = ".tif")
download.file(url = "https://github.com/STBrinkmann/data/raw/main/GVI_Data/GVI_DSM.tif",
destfile = DSM_tmp, mode="wb")
# Download GreenSpace
GS_tmp <- tempfile(fileext = ".tif")
download.file(url = "https://github.com/STBrinkmann/data/raw/main/GVI_Data/GVI_GreenSpace.tif",
destfile = GS_tmp, mode="wb")
```
Load DSM, DEM and Greenspace Mask, and generate the observer location as a `sf` `POINT` feature.
```{r message=FALSE, warning=FALSE}
# Load libraries. if one is not installed, use the "install.packages()" function
library(terra)
library(sf)
library(sfheaders)
# Load raster objects
DEM <- rast(DEM_tmp)
DSM <- rast(DSM_tmp)
GreenSpace <- rast(GS_tmp)
# Generate single observer point
observer <- st_sf(sf_point(c(492243.3, 5454231.4)), crs = st_crs(26910))
```
### 1. Single Point
Calculate the viewshed for a 200 meters radius around the observers position at 1.7 meters height (eye level).
```{r message=FALSE, warning=FALSE, fig.height=5, fig.width=10}
library(GVI)
vs <- GVI::viewshed(observer = observer, dsm_rast = DSM, dtm_rast = DEM,
max_distance = 200, observer_height = 1.7, plot = TRUE)
```
*Left: Digital Surface Model (DSM); Right: Viewshed, where green = visible and yellow = no-visible area.*
The Viewshed Greenness Visibility Index (VGVI) can now be calculated using the `vgvi` function.
```{r}
vgvi_from_sf(observer = observer,
dsm_rast = DSM, dtm_rast = DEM, greenspace_rast = GreenSpace,
max_distance = 200, observer_height = 1.7,
m = 0.5, b = 8, mode = "logit")$VGVI
```
The output of \~0.525 indicates, that \~53.5% of the visible area, within a 200 meters radius, is green.
### 2. Road Network
We also provide sample data of a sf line feature to demonstrate the `vgvi_from_sf` function. This feature represents roads and paths, that can be reached by walking within 5 minutes from our observer location.
```{r}
# Download line feature and load as sf object
isodistance <- read_sf("https://github.com/STBrinkmann/data/raw/main/GVI_Data/isoline.gpkg")
```
Compute the VGVI along the line feature. Since the resolution is 5 meters, points along the line feature will be generated every 5 meters.
```{r}
vgvi_sf <- vgvi_from_sf(observer = isodistance,
dsm_rast = DSM, dtm_rast = DEM, greenspace_rast = GreenSpace,
max_distance = 200, observer_height = 1.7,
m = 1, b = 3, mode = "exponential", cores = 12)
```
 *Output of the* `vgvi_from_sf` *function. Red points indicate low, green points indicate high values of VGVI. Paths and roads along the park tend to have high visible greenness, the road in the east has low visible greenness.*
```{r echo=FALSE}
# Delete raw DTM, DSM and GreenSpace files
rm(DSM, DEM, GreenSpace)
unlink(c(DSM_tmp, DEM_tmp, GS_tmp))
```
### 3. Large Area of Interest
Often researchers need to compute the VGVI for a large study area, therefore we have provided a sample workflow for the City of Vancouver. Since this workflow is extensive, we have uploaded it on an [external website](https://geobrinkmann.com/post/visible-greenness-exposure-index-an-example-workflow-for-the-city-of-vancouver/). Sample Datasets for Greater Manchester and Vancouver, BC can be downloaded from [Zenodo](https://doi.org/10.5281/zenodo.5061256).
## About
### Citation {#citation}
Run this command to get info on how to cite this package.
```{r warning = FALSE}
citation("GVI")
```
### Package contributors
Brinkmann, Sebastian (Creator and author) e-mail: [sebastian.brinkmann\@fau.de](mailto:[email protected])
Dr. S.M. Labib (Author) e-mail: [sml80\@medschl.cam.ac.uk](mailto:[email protected])
### Paper authors
S.M. Labib (1, 2\*)\
Jonny J. Huck (1)\
Sarah Lindley (1)
1: Department of Geography, School of Environment, Education and Development (SEED), University of Manchester, Arthur Lewis building (1st Floor), Oxford Road, Manchester M13 9PL, United Kingdom.
2: Centre for Diet and Activity Research (CEDAR), MRC Epidemiology Unit, University of Cambridge, Clifford Allbutt Building, CB2 0AH, Cambridge, United Kingdom.
\*corresponding author
## Bibliography
Bresenham, J.E. (1965): Algorithm for computer control of a digital plotter. IBM Systems Journal, vol. 4, no. 1, pp. 25-30, doi: [10.1147/sj.41.0025](https://doi.org/10.1147/sj.41.0025).
Bresenham, Jack (1977): A linear algorithm for incremental digital display of circular arcs. Commun. ACM 20, 2 (Feb. 1977), 100--106. doi: [10.1145/359423.359432](https://doi.org/10.1145/359423.359432).
Brinkmann, S. T., Kremer, D., and Walker, B. B. (2022): Modelling eye-level visibility of urban green space: Optimising city-wide point-based viewshed computations through prototyping, AGILE GIScience Ser., 3, 27, [https://doi.org/10.5194/agile-giss-3-27-2022](https://doi.org/10.5194/agile-giss-3-27-2022,).
Labib, S.M., Jonny J. Huck, and Sarah Lindley (2021): Modelling and Mapping Eye-Level Greenness Visibility Exposure Using Multi-Source Data at High Spatial Resolutions. *Science of The Total Environment* 755 (February): 143050. doi: [j.scitotenv.2020.143050](https://doi.org/10.1016/j.scitotenv.2020.143050).