A low-cost and low-weight sensor system for urban heat island surveys and education

Abstract

We developed “Meteobike”, a light and low-cost urban traverse system based on the Raspberry-Pi. It can be assembled and programmed in undergraduate classes. The system can be mounted on bicycles to record geotagged air temperatures and humidity as one moves through cities and adjacent rural areas. In summer 2018, we used 25 systems to map and analyze the detailed spatial patterns of Freiburg’s canopy layer urban heat island.

Full text

Surveying Freiburg’s CLUHI with bikes
Students built a fleet of 25 “Meteo-
bike” sensors in our undergraduate
course on meteorological instrumen-
tation. Prior to the traverses, we cali-
brated sensors against a reference
CS215 (Campbell Scientific Inc.) and
we applied sensor specific corrections to reduce systematic
biases in temperature and humidity readings. With the sys-
tems running over 2 hours on 25 bikes, we were able to cap-
ture the detailed patterns of nocturnal air temperatures.
The canopy layer urban heat island (CLUHI) is the most
widely studied urban climate phenomenon. In particular, the
intra-urban variability of the CLUHI has lately received sig-
nificant attention to assess how climate change is increasing
the risk of heat stress in German cities. Hence, the CLUHI is
an excellent topic to stimulate undergraduate engagement
in environmental meteorology, combining aspects of atmo-
spheric observation technology and geospatial analysis in
the context of a changing climate.
Design of the Meteobike system
We developped a new urban heat island traversing system
based on the Raspberry Pi Zero W microcontroller (Rasp-
berry Foundation) to collect data from digital low-cost ther-
mometers / hygrometers (DHT22 / AM2302) and merges
this data with GPS location, altitude and speed recorded by a
GPS (Adafruit Ultimate GPS Breakout V3).
Temperature and humidity sensors are housed in a reflective
tube to provide a basic radiation shield. We found that if cy-
cling speed is ensured at > 8 km/h, sensors are suciently
ventilated.
A low-cost and low-weight sensor system
for urban heat island surveys and education
We developed “Meteobike”, a light and low-cost
urban traverse system based on the Raspberry-Pi.
It can be assembled and programmed in
undergraduate classes. The system can be
mounted on bicycles to record geotagged air
temperatures
and humidity
as one moves
through cities
and adjacent rural
areas. In summer 2018,
we used 25 systems
to map and analyze the
detailed spatial patterns
of Freiburg’s canopy
layer urban heat
island.
Andreas Christen(1), Felix Baab(1), Heinz Christen(2), Amy Laframboise(1,3), Joey K. Lee(4), Natasha Picone(5), and Dirk Redepenning(1)
Acknowledgements - This project was funded through funding for teaching and start-up at the Uni-
versity of Freiburg. We acknowledge the enthousiastic support by our 2nd year undergraduate students
“Environmental Meteorology”. Students Nora Bergener and Yuting Wu contributed with data.
Selected photos are provided by the University of Freiburg Public Relations oce.
Aliations and contacts - (1) Chair of Environmental Meteorology, Faculty of Environment and Natu-
ral Resources, University of Freiburg, Germany - (2) IT Services, University of Basel, Basel, Switzerland
(retired) - (3) Environmental Science, Laurentian University, Greater Sudbury, ON, Canada - (4) New Yok
University, New York, NY, USA - (5) CIG - IGEHCS - CONICET/UNCPBA Tandil, Argentina. Primary au-
thor contact email: andreas.christen@meteo.uni-freiburg.de
A fine-resolution analysis of Freiburg’s CLUHI
In our course, students explored the influence of urban den-
sity, green fraction, water bodies or orography on nocturnal
temperatures. Measurements were gridded and merged with
other geospatial datasets for a quantitative analysis.
Routes covered the entire city, reaching out into valleys and
rural areas of the Upper Rhine Plain. Within two hours, stu-
dents biked collectively more than 400 km and collected
10,000+ datapoints per evening. Data are expressed as the
dierence between values measured on bikes and simulta-
neously measured values at a station in the city centre.
The ingredients for a
“Meteobike” are simple
and aordable - total cost
per system is about €60
(excluding the smart-
phone). Systems are in-
dependently powered us-
ing a cell-phone battery
bank. All components,
including the smartphone
are placed in a simple ac-
cessory bag that can be
attached to any bicycle.
Freiburg’s urban climate is interesting and complex, as thermo-topgraphic
winds, specifically drainage flows and the cool mountain wind “Höllentäler”
along the Dreisamtal (Photo), lower the intensity of the nocturnal canopy
layer heat island.
Visualization of the CLUHI at
70 x 70 m resolution in Google
Earth. Here areas in red are
warmer, areas in blue are cooler.
The gridded dataset allowed for
detailed georeferencing and cal-
culation of area-averages.
A comparison of area-averaged temperatures attributed to dierent low-
density local climate zones (LCZs) in the outskirts to the East (blue, valleys,
hillsides) and the West (orange, Upper Rhine Valley plain) of the city. In
nearly all cases, for same density, the western outskirts are warmer - here
the influence of the drainage and mountain winds is reduced or absent.
This project is open-source, instructions and
code can be freely downloaded at
https://github.com/achristen/Meteobike
Web-based visualization of temperature dierences of all traces during the
night of June 19, 2018. Note the lower temperatures in valleys to the East
and South, explained by cold air drainage and/or mountain winds.
Conclusions
This project presented a rewarding first opportunity for un-
dergraduate students to engage in data collection and data
analysis techniques using web-based tools. The project intro-
duced students to the challenges of crowd-sourced data and
geospatial data visualization, and engaged them in approach-
ing and addressing environmental meteorological problems.
LCZ 5
LCZ 6
LCZ 9
LCZ A
LCZ B
LCZ D
Air temperature di erence to city centre (K)
-4
-3
-2
-1
0
-2,8
-1,0
-3,9
-0,7
-0,6
-0,5
-3,8
-2,4
-3,0
-2,3
-1,7
-1,7
Freiburg East
Freiburg West