Public Lab Research note

Collecting data about Stormwater Issues through Citizen Science

by panicpuri | September 12, 2019 13:57 12 Sep 13:57 | #20842 | #20842


I am a PhD student trying to build community resilience to flooding events by engaging more citizens in the data collection, processing, and analysis pipeline. I am working with the Nine Mile Run Watershed Association (NMRWA) (non-profit) to collect data that will prove the efficacy of installed Green Stormwater Infrastructure (GSI) in reducing flooding and runoff. I hope to conduct a workshop for interested citizens to build and install their own DIY sensor systems using Arduino boards and sensors in their backyards to measure temperature, air quality, soil moisture, soil pH, and precipitation. I also hope to study the different ways in which this data can be communicated to a web server for visualization - WiFi, GSM (SIM card), LoRaWAN (Low power, long range). And finally, I want to see the engagement and willingness of those who signed up to follow through with taking measurements with the assistance of prompts and reminders. Ideally, the Arduino sensor system will be install-and-forget, as it sends a stream of data periodically and is in a low-power state in-between sensing.

Our main concern:

Our main concern is the workshop for citizens to make their own sensor systems - and the initial installation in their backyards. The Arduino is not 'too' complicated and I have worked with Raspberry Pis and an Arduino before, but I am wary that the workshop might be throwing not-so-tech-savvy folks under the bus.

Obstacles and supporting information:

As of now, there are several spots in the watershed where NMRWA has installed Green Stormwater Infrastructure and introduced tree-planting efforts. However, the challenge comes in quantifying the benefit that these initiatives have had on the local community and ecosystem.

We hope to be able to point to differences in data between canopied areas and those without trees in terms of temperature (urban heat island effect) and air quality; and we hope to use soil data to investigate the differences in areas that have GSI close-by and those that don't.

We are trying to do this as economically as possible, trying to keep our sensor system affordable (~$50), although my school will be funding the first round of 10 sensor systems. I have built a mobile weather station with solar charging before that cost ~$70, and I do believe a $50 sensing system is possible. I have tallied the costs of individual sensors, systems, and peripheral components, and if we order the components from aliexpress instead of amazon, the system could be built for ~$33!

I have not worked with soil sensing or air quality sensing before, and I was hoping to get some suggestions or advice from this community about which sensors/systems work best for this type of project. Also open to any criticism, recommendations, links, etc. that will make this project successful.

Who is engaged in this concern?

The NMRWA has been closely engaged with local community members on this issue and have been working towards more green infrastructure and watershed management strategies that work for everyone. My advisor and I are now getting involved as well, to help use our skillset to gain new insights about how/where/why GSI is helping the community, and how the community can return the favour.

What are the initial questions?

Are there any sensors or systems that might be more efficient/inexpensive than these?

Is there a good resource for building a sensor system pretty close to this? (I looked, and found a few similar ones, but nothing that combined soil, air, and temp sensing)

How do I make sure this stays waterproof?

I plan to power the system with a 9v rechargeable battery (link in the doc), approx how long will this last given a sensing interval of 1hr?

What are some possible problems I could run into getting this project off the ground?

I have been planning this project for about a week, and I will be implementing it in the coming months. I would greatly appreciate any input from this community on how to more meaningfully engage citizens and make this experience as seamless as possible. I also hope to be able to write a paper about the challenges faced while setting up this system - as I predict there will be (at least) a few. I am a firm believer in open source principles and I am hoping to build something that can be replicated in other places and engage many more people so that they are aware of the issues that affect their community.

Thank you for creating this platform and allowing this exchange of ideas!

Title Author Updated Likes Comments
Nothing yet on the topic "stormy-pgh" -- be the first to post something!


It looks like you have made a really good start with this project. I agree that a challenge will be to avoid scaring away the neighborhood residents with unwieldy electronics. One approach might be to start small with a simple device that meets a subset of your goals.

For example, if you wanted to start with a system that:
1. had to be plugged into house power (extension cord into a backyard)
2. measured only air temperature, soil temperature, humidity, barometric pressure, light level, and maybe a few others
3. sent its data to the web via home Wi-Fi

this could be done for less than $10.00. See this note for the basic approach using an ESP8266. This approach could be especially attractive to users with smart phones who could see their backyard data in real time on a phone app.

If it is critical that the devices be powered by battery, the above approach will have to be modified so that the device either turns off between data readings or is put into a sleep state. This is probably possible, but I have not figured that part out yet with the above hardware. So another approach is a device which saves data to SD card (no web connection) but can run on batteries for a year. This device is based on an Arduino Pro Mini and TPL5110 timer.

Achieving your goal of a device which can last for weeks or months on batteries, sends its data to the web, and can run a Plantower dust sensor (which needs 5 volts) requires a more sophisticated solution. Such solutions are available but add a new level of complexity.


Thank you for your response, @cfastie ! The SensorShip looks like a really interesting device and I will definitely consider using that in a part of the project. I have used a BME280 sensor for a weather station project before, but need to incorporate the soil and air quality sensing measurements. I am just not sure about the power requirements of the different sensors and whether the smaller microcontroller will be able to operate all my sensors.

The Mini Pearl Snap also looks like it could fulfill my requirements and part of the reason I posted here was so that I didn't try reinventing the wheel while designing this system. Thank you so so much for linking to similar projects!

I am still in the design stages, so this is extremely helpful.

I think you could find residents who want to participate, and could double as field inspectors for your units, if you got them on board to install units on their property. People like to know what's going on, and they like discussing things with their neighbors.

Perhaps that's easy for me to say, since i work not for the university, but for a membership based organization where the dues paying members influence the direction of projects. I can see where there's often tension between a large, land-owning institution that may have been a community bully in the past, and residents, who are often very interested in the local data, and having someone talk to them about what they are seeing, if we can get past the class politics.

Reply to this comment...

We shared this on Public Lab's social media. The Facebook post received this reply: Is there already a community of folks interested in collecting data? Also, schools are, generally speaking, really into technology. It is fairly common for a school to have a maker space. They take many forms. Maybe reach out to a local school/ district to introduce project and offer the opportunity to build the sensors?

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

How are things advancing? Quite well, i hope!

Is this a question? Click here to post it to the Questions page.

Things are advancing (although a bit slower than I hoped, but I guess that's to be expected). I am still designing a plan for the entire system (including community input, hardware, DIY workshop, IRB approval, metrics to measure). In the meanwhile, I'm waiting for my components to arrive so I can start prototyping, and trying to model the system using NetLogo (Agent-based-modelling).

We have agreed on a timeline where we model and prototype until the end of this year (2019) to make sure we have a rugged, accurate system - and the DIY workshop some time in January or February 2020. Partly because obtaining IRB approval is a time-consuming process, partly because we just want to be thorough and not sink a lot of money on faulty equipment/ bad system design. The people I'm working with at the NGO assured me that it's better to spend a little bit longer on planning than spend a lot more time troubleshooting issues in the middle of the project implementation.

I will post my models and research here in the meanwhile, as updates to this project! Thank you so much for your input and engagement. 😃

wow great to hear and thanks for the timeline update! We're cheering for you and the team. What system are you thinking about for "prompts and reminders" for the people collecting data?

Is this a question? Click here to post it to the Questions page.

I have worked with If-This-Then-That (IFTTT)* systems in the past, and they were simple, yet effective. A system that routinely sends text message reminders might be useful, or even a push-notification app (although I don't think people want to install yet another app).

My advisor and I are also talking about social-sensing, something along the lines of a Facebook group (although more inclusive - and cognizant of the droves of people quitting Facebook) where the sensor-systems can post updates when something is wrong (ex: "Arrr-duino: I think my temperature sensor has failed. Could someone help figure out what's wrong?"), or when some human intervention is necessary, or even simply to convey a fun-fact (ex: "The air quality has been in the normal range all month! Much better than last month where 10/31 days saw poor air quality!"). This way, the community can be engaged and aware, and provide a support system, with more tech-minded folks helping out the ones who don't have the time/technical expertise to troubleshoot.

Is this a question? Click here to post it to the Questions page.

Reply to this comment...

Login to comment.