OCTOPUS v.2 Web Feature Service (WFS) in R

User guide and brief demonstration of capabilities

Preface

This document describes how to use OCTOPUS WFS through R software ¹ environment – a free, open-source, reproducible and highly versatile way of data access.

OCTOPUS database

OCTOPUS (https://octopusdata.org) is a geospatial open access database of sedimentological, archaeological, and paleo-environmental ages and process rate determinations, organised into thematic data collections. These data can be viewed and exported through a bespoke web interface ² or accessed directly. For the latter, OCTOPUS database provides OGC compliant web feature service (WFS) allowing users to access database content via third party software such as Geographic Information Systems (GIS) ³, or R language. OCTOPUS data is served by GeoServer ⁴, an open server solution for sharing geospatial data. Unless data from OCTOPUS is explicitly downloaded and locally stored by the user, it will remain cloud-borne ⁵, so the user will not have to care about data storage and being up-to-date.

WFS in a nutshell

Geospatial and location information development and standardisation globally is overseen by the Open Geospatial Consortium (OGC) ⁶. Web Feature Service (WFS) is an OGC interface standard that enables platform independent requests for spatial features across the internet. This is accomplished by Geography Markup Language (GML), an XML derivative. Unlike for WMS (Web Map Service), where immutable map tiles are returned, WFS vector entities can be queried, altered, and spatially analysed.
WFS functionality knows three basal operations - GetCapabilities, DescribeFeatureType, and GetFeature. Calling GetCapabilities will generate a standardised, human readable meta-dataset that describes a WFS service and its functionality. DescribeFeatureType produces an overview of supported feature types, and GetFeature fetches features including their geometry and attribite values, i.e, variable fields.

OCTOPUS WFS data access via R/RStudio

The below demo R script fetches, via WFS, spatial layers including rich attribute data from OCTOPUS database and generates a scatter plot (Plot 1) and an interactive map representation (Plot 2), respectively.

Note - The script requires the below packages. If not installed on your machine yet, run

# install.packages(c("sf","httr","tidyverse","ows4R","viridis", "mapview", dependencies = TRUE))

and you’ll be all set up.

First we’re going to load the required packages

library(sf) # Simple features support (sf = standardized way to encode spatial vector data)
library(httr) # Generic web-service package for working with HTTP
require(tidyverse) # Workhorse collection of R packages for data sciences
library(ows4R) # Interface for OGC web-services incl. WFS
library(viridis) # Predefined colorblind-friendly color scales for R

OK, we’re ready to go now.
In the following we store the OCTOPUS WFS URL in an object.
Then, using the latter, we establish a connection to OCTOPUS database.

OCTOPUSdata <- "http://geoserver.octopusdata.org/geoserver/wfs" # store url in object
OCTOPUSdata_client <- WFSClient$new(OCTOPUSdata, serviceVersion = "2.0.0") # connection to db

Let’s see what is there, i.e. show all available layer names and titles

OCTOPUSdata_client$getFeatureTypes(pretty = TRUE) # show available layers and titles

##                                            name
## 1                    be10-denude:crn_aus_basins
## 2                   be10-denude:crn_aus_outlets
## 3                    be10-denude:crn_int_basins
## 4                   be10-denude:crn_int_outlets
## 5                    be10-denude:crn_xxl_basins
## 6                   be10-denude:crn_xxl_outlets
## 7                 be10-denude:crn_inprep_basins
## 8                be10-denude:crn_inprep_outlets
## 9                      be10-denude:publications
## 10                            opengeo:countries
## 11                           be10-denude:expage
## 12 be10-denude:fossahul_webmercator_nrand_25000
## 13                    be10-denude:sahularch_osl
## 14                    be10-denude:sahularch_c14
## 15                     be10-denude:sahularch_tl
## 16             be10-denude:sahulsed_aeolian_osl
## 17              be10-denude:sahulsed_aeolian_tl
## 18             be10-denude:sahulsed_fluvial_osl
## 19              be10-denude:sahulsed_fluvial_tl
## 20          be10-denude:sahulsed_lacustrine_osl
## 21           be10-denude:sahulsed_lacustrine_tl
##                                         title
## 1     CRN Australian collection: River basins
## 2     CRN Australian collection: Sample sites
## 3         CRN Global collection: River basins
## 4         CRN Global collection: Sample sites
## 5              CRN Large basins: River basins
## 6              CRN Large basins: Sample sites
## 7      CRN UOW (in preparation): River basins
## 8      CRN UOW (in preparation): Sample sites
## 9                    CRN basin bounding boxes
## 10                     Countries of the World
## 11                            ExpAge Database
## 12                          FosSahul Database
## 13          Sahul Archaeology: OSL collection
## 14  Sahul Archaeology: Radiocarbon collection
## 15           Sahul Archaeology: TL collection
## 16    Sahul Sedimentary Archives: Aeolian OSL
## 17     Sahul Sedimentary Archives: Aeolian TL
## 18    Sahul Sedimentary Archives: Fluvial OSL
## 19     Sahul Sedimentary Archives: Fluvial TL
## 20 Sahul Sedimentary Archives: Lacustrine OSL
## 21  Sahul Sedimentary Archives: Lacustrine TL

That’s basically it. Talking to the database via WFS takes three short lines of code. Everything below this line does not deal with data access anymore, but with data presentation.
BTW A full description of OCTOPUS database and its collections can be found in a dedicated Earth Systems Science Data publication.

Example 1. Australian 10Be-derived catchment-averaged denudation rates

Here we fetch and plot Australian catchment-averaged 10Be denudation rates (i.e., layer ‘be10-denude:crn_aus_basins’ from the above list)

url <- parse_url(OCTOPUSdata) # parse URL into list
url$query <- list(service = "wfs",
                  version = "2.0.0",
                  request = "GetFeature",
                  typename = "be10-denude:crn_aus_basins",
                  srsName = "EPSG:900913") # set parameters for url$query

request <- build_url(url) # build a request URL from 'url' list
CRN_AUS_basins <- read_sf(request) # read simple features using 'request' URL. Takes few secs...

Now that we have the data available, we define our plot parameters. We want to plot denudation rate (“EBE_MMKYR”) over average slope gradient (“SLP_AVE”) and call the plot (last line)

myPlot <- ggplot(CRN_AUS_basins, aes(x=SLP_AVE, y=EBE_MMKYR)) + # plot denudation rate over average slope
geom_errorbar(aes(ymin=(EBE_MMKYR-EBE_ERR), ymax=(EBE_MMKYR+EBE_ERR)), linewidth=0.3, colour="gray80") + # visualise errors
geom_point(aes(size=AREA, color=ELEV_AVE), alpha=.5) + # scale pts. to "AREA", colour pts. to "ELEV_AVE"
scale_color_viridis(option="C", direction = -1) + # use 'viridis' colour scale
scale_size_continuous(range = c(2, 10)) + # define point size range for better visibility
xlab("Slope gradient [m km^-1]") + ylab("Denudation rate [mm kyr^-1]") + # set labels for x and y axes
ggtitle("Australian 10Be catchment-avg. denudation rates") + # make title
theme(plot.title = element_text(size = 18, face = "bold")) + # title settings
labs(size = "Catchment \narea [km^2]", colour = "Average \ncatchment \nelevation [m]") # re-label legend
myPlot # make plot

Plot 1. Australian 10Be-derived catchment-averaged denudation rates against average slope gradient. For more detail see CRN/Australia collection info (https://doi.org/10.4225/48/5a836cdfac9b5)

Example 2. Australian sedimentary fluvial OSL ages

For this example we quickly want to display Australian OSL (Optically Stimulated Luminescence) ages on a base map.

library(mapview) # Provides functions for quick visualisation of spatial data
mapviewOptions(fgb = FALSE)
url2 <- parse_url(OCTOPUSdata) # parse URL into list
url2$query <- list(service = "wfs",
                   version = "2.0.0",
                   request = "GetFeature",
                   typename = "be10-denude:sahulsed_fluvial_osl",
                   srsName = "EPSG:900913") # set parameters for url$query

request2 <- build_url(url2) # build request URL from 'url' list
SahulSed.FLV.OSL <- read_sf(request2) # read simple features using 'request' URL. Takes few secs...
SahulSed.FLV.OSL <- st_set_crs(SahulSed.FLV.OSL, 900913) # Set Coordinate Reference System
SahulSed.FLV.OSL = st_transform(SahulSed.FLV.OSL,
        crs = "+proj=longlat +datum=WGS84") # Transform geometry to geographic coordinates, WGS84
mapview(SahulSed.FLV.OSL, xcol = "X_WGS84", ycol = "Y_WGS84",
        zcol = "OSL_AGE", at = seq(0, 350, 50), alpha = .25, # set range (0 to 350 ka) and bins (50 ka)
        alpha.regions = 0.1, legend = TRUE) # Display on map using "mapview" package

Plot 2. Australian sedimentary fluvial OSL ages (ka). For more detail see SahulSed/fluvial/osl collection info (https://doi.org/10.25900/p5ye-rn35)

All done!

Note that, thanks to the very slick ‘Mapview’ ⁷ functionality, points on the map are mouse-over sensitive and can be queried in depth by clicking. Further, the map is scalable and you can choose between a decent selection of base map layers.

---

PS For more information about OCTOPUS database go and see the OCTOPUS database documentation, OCTOPUS visual database schema, OCTOPUS GitHub repositories, OCTOPUS resources on CABAH website, and Zenodo OCTOPUS community.

Acknowledgement. OCTOPUS database is supported by the Australian Research Council Centre of Excellence for Australian Biodiversity and Heritage (CABAH) ⁸. Funding was also provided by the University of Wollongong ⁹ and the Australian National Data Service (ANDS), now part of the Australian Research Data Commons ¹⁰.

---

1. https://www.r-project.org/↩︎

2. http://octopusdata.org↩︎

3. A step-by-step description can be found in OCTOPUS database documentation https://octopus-database.readthedocs.io/en/latest/usage.html#wfs-data-access-via-qgis↩︎

4. http://geoserver.octopusdata.org↩︎

5. OCTOPUS is implemented in PostgreSQL (https://www.postgresql.org) and sits on Google Cloud Platform (GCP)↩︎

6. https://www.ogc.org↩︎

7. https://r-spatial.github.io/mapview/↩︎

8. https://epicaustralia.org.au↩︎

9. https://www.uow.edu.au↩︎

10. https://ardc.edu.au↩︎