High-Level Project Summary
The scope of this project, Hebybu! (Helped by bugs!) is to use insects as bioindicators for illegal pollution - i.e., biological sensors (think of a canary in a coal mine!) – fusing Earth Observation data and citizen science. Observations of insects on the ground could lead to the identification of extended areas containing illegally dumped waste.
Link to Project "Demo"
Link to Final Project
Detailed Project Description
Without insects, life on our planet would hardly exist in the way we currently know.
Like the rain, which allows plants to nourish themselves in places far from the edge of bodies of water, insects are responsible for capillary transport phenomena (e.g., flower-to-flower) and transformation processes, promoting growth and resilience for many other species.
However, while acquiring the resources needed for their sustenance, insects can be harmful to humans, e.g., agricultural pests, disease-transmitting bugs, or termites chewing on your favourite bookshelf. The core importance of insects in the anthroposphere has led to many studies on how to track these creatures, using tools ranging from ground stations up to orbital observations.
A tool that is proving extremely successful and useful to generate large scale observations and classifications on the ground is citizen science. Many citizen science projects have already been successful in targeting insects – as for example Caterpillars Count! (https://caterpillarscount.unc.edu/) or the app Cicada Safari (https://cicadasafari.org/).
This emerging capability can be exploited for ubiquitous, all-weather, ground monitoring, and coupled with the wide availability of spaceborne observation platforms. Observing insects from orbit is already within the realm of possibility: examples are the work of Bhattarai et al. [1], investigating the detection of Hessian flies using the NDVI, or Hollaus and Vreugdenhil [2] who have analysed the use of radar payloads to analyse the presence of bark beetles. This fusion of local and Earth Observation (EO) data would allow to monitor, link and merge trends and events at multiple scales in space and time.
Interestingly, the previously mentioned bark beetles have been shown to interact with wildfires, further proving that insects strongly interact with large-scale environmental phenomena. This project aims to exploit similar interactions of this kind and blend people’s intuition and processing power over a large amount of remote sensing data. An application seems to be of immediate interest for this approach: monitoring illegal pollution. For example, a particular pollutant might increase the rate of reproduction of an insect laying its eggs in the water, leading to a change in the surface properties of bodies of water. The change in electromagnetic properties could then be detected by satellites, which would be able to quantify the extent of the environmental crime.
Therefore, in short: the detection of insects on the ground by human triggers intense EO data acquisitions targeting the insects themselves to identify illegal wastes of dump.
For each sensor available on a satellite, this is achieved by defining a set of high visibility species (HVS) susceptible to the pollutant of interest and generating a strong response in the sensor. An HVS might be as such for a sensor because of physical properties (e.g. a high response in the band), social behaviours (e.g. large scale swarms) or behaviours with affect the landscape. This might even be some reactive change to an environment becoming overly harsh.
Specifically, Hebybu! capitalises on the following workflow:
•People can download an app or login to a website to report in real time the presence of a particular insect
•A large or anomalous activity from a certain species can trigger a wide-area EO survey of the interested location, and in general be used to train machine learning models to connect remote sensing and local phenomena
•The calibrated EO data allow to identify the source and extent of the pollutants, helping in either stopping the illegal activity or limiting its effects
Space Agency Data
MODIS, VIIRS, GIBS data, population distribution models, biological models, NDVI & EVI, Lidar data, Sentinel radar data,
Hackathon Journey
Funnily enough, the inspiration for this challenge came from a paper I wrote a couple of years ago, titled "Monitoring Dust Motion Around Airless Celestial Bodies: Characterizing Suitable Landing Zones". On an asteroid, winnowed dust, over long timescales, should intrinsically converge towards zones of interest for a landing probe.
If geological processes can be thought of as unconventional computational processes returning data of interest for remote sensing, the same principle can obviously be extended to biology, as it has been proven over and over that biological processes can perform extremely elegant computations (e.g. https://watersource.awa.asn.au/technology/innovation/polish-city-using-mussels-monitor-water-quality/, https://www.wired.com/2010/01/slime-mold-grows-network-just-like-tokyo-rail-system/).
As usual I would like to thank the organisers for giving a platform to be creative and explore fascinating, and I would like to thank Claire for being the most amazing and inspiring human being I've ever met.
References
[1] Bhattarai, G. P., Schmid, R. B. & McCornack, B. P. Remote Sensing Data to Detect Hessian Fly Infestation in Commercial Wheat Fields. Sci. Reports 2019 919, 1–8 (2019).
[2] Hollaus, M. & Vreugdenhil, M. Radar Satellite Imagery for Detecting Bark Beetle Outbreaks in Forests. Curr. For. Reports 2019 545, 240–250 (2019).
Global Judging
This project has been submitted for consideration during the Judging process.