Exploring the openSenseMap Dataset

knitr::opts_chunk$set(echo = TRUE)

This package provides data ingestion functions for almost any data stored on the open data platform for environemental sensordata https://opensensemap.org. Its main goals are to provide means for:

Exploring the dataset

Before we look at actual observations, lets get a grasp of the openSenseMap datasets' structure.


all_sensors = osem_boxes()

This gives a good overview already: As of writing this, there are more than 700 sensor stations, of which ~50% are currently running. Most of them are placed outdoors and have around 5 sensors each. The oldest station is from May 2014, while the latest station was registered a couple of minutes ago.

Another feature of interest is the spatial distribution of the boxes: plot() can help us out here. This function requires a bunch of optional dependencies though.

if (!require('maps'))     install.packages('maps')
if (!require('maptools')) install.packages('maptools')
if (!require('rgeos'))    install.packages('rgeos')


It seems we have to reduce our area of interest to Germany.

But what do these sensor stations actually measure? Lets find out. osem_phenomena() gives us a named list of of the counts of each observed phenomenon for the given set of sensor stations:

phenoms = osem_phenomena(all_sensors)

Thats quite some noise there, with many phenomena being measured by a single sensor only, or many duplicated phenomena due to slightly different spellings. We should clean that up, but for now let's just filter out the noise and find those phenomena with high sensor numbers:

phenoms[phenoms > 20]

Alright, temperature it is! Fine particulate matter (PM2.5) seems to be more interesting to analyze though. We should check how many sensor stations provide useful data: We want only those boxes with a PM2.5 sensor, that are placed outdoors and are currently submitting measurements:

pm25_sensors = osem_boxes(
  exposure = 'outdoor',
  date = Sys.time(), # ±4 hours
  phenomenon = 'PM2.5'

Thats still more than 200 measuring stations, we can work with that.

Analyzing sensor data

Having analyzed the available data sources, let's finally get some measurements. We could call osem_measurements(pm25_sensors) now, however we are focusing on a restricted area of interest, the city of Berlin. Luckily we can get the measurements filtered by a bounding box:


# construct a bounding box: 12 kilometers around Berlin
berlin = st_point(c(13.4034, 52.5120)) %>%
  st_sfc(crs = 4326) %>%
  st_transform(3857) %>% # allow setting a buffer in meters
  st_buffer(set_units(12, km)) %>%
  st_transform(4326) %>% # the opensensemap expects WGS 84
pm25 = osem_measurements(
  phenomenon = 'PM2.5',
  from = now() - days(3), # defaults to 2 days
  to = now()


Now we can get started with actual spatiotemporal data analysis. First, lets mask the seemingly uncalibrated sensors:

outliers = filter(pm25, value > 100)$sensorId
bad_sensors = outliers[, drop = T] %>% levels()

pm25 = mutate(pm25, invalid = sensorId %in% bad_sensors)

Then plot the measuring locations, flagging the outliers:

st_as_sf(pm25) %>% st_geometry() %>% plot(col = factor(pm25$invalid), axes = T)

Removing these sensors yields a nicer time series plot:

pm25 %>% filter(invalid == FALSE) %>% plot()

Further analysis: comparison with LANUV data TODO

Try the opensensmapr package in your browser

Any scripts or data that you put into this service are public.

opensensmapr documentation built on May 2, 2019, 2:58 p.m.