Part 2: Making sense of sensors in lakes
Missing the lake for the water
This is the second blogpost in the Lake Sensors series where we trace the journey of CSEI’s Urban Lakes Initiative. If you would like to collaborate with us on lakes, email us at firstname.lastname@example.org. Read Part 1: Making sense of sensors in lakes before reading further.
Phase 2: The pivot to citizen science
The failure of our IoT venture did not discourage us from our quest to improve data for better management of lakes. We pivoted to a ‘humans as sensors’ approach — we would collaborate with lake groups and low-cost sensing technology partners and encourage citizens lake group volunteers to measure water quality in lakes. We partnered with the Foundation for Environmental Monitoring (FFEM), a local Bengaluru-based start-up that had developed a smartphone-based test kit designed for monitoring water in lakes and rivers. The kit was able to measure nitrates and phosphates. (We also tested chlorophyll and Dissolved Oxygen kits but these were not accurate.)
But getting humans to collect data is not an easy task. There are several challenges to creating a motivated cadre of people willing to venture out at all hours of the day to collect samples.
What we learned from the Citizen Science Phase:
First, technology needs to be intuitive, simple, and stable. We ran several training programmes. One early problem was that people had different smartphones. The FFEM smartphone test was based on a ‘colorimetric’ method. It involved adding a few ml of a reagent (chemical) to the water sample. This triggered a chemical reaction, which would change the colour of the water depending on how much of the pollutant of interest there was in the water. The smartphones would essentially take a picture of the coloured water and compare it to a pre-defined colour chart that linked the brightness of the colour to a pollutant concentration.
Because all the smartphones had slightly different cameras, each had to be calibrated individually and this was a very time-consuming process. Additionally, every time a citizen scientist changed their phone, the whole process would have to be repeated. The problem has since been rectified.
Second, people need to see a clear benefit from their participation. Much has been written about the value of citizen collected data for science. But what we didn't realize is that, unlike birding or wildlife observation data, what we were really asking citizens to do was get up before sunrise and wade into a stinky lake; this was not an inherently rewarding activity.
Therefore, citizens needed a strong motivation or at least a theory of change to be convinced that the effort was worthwhile. But we were not always able to provide this. Knowing a lake is polluted didn't always translate immediately to solutions. Because Bengaluru’s lakes are connected, often the pollutants come from upstream. Some of the problems like raw sewage required large infrastructure investments, which also take time to build. So there was very little an individual lake group could do about it. Smaller-scale solutions that are within the reach of citizen groups are often untested or only partially effective. This led to a lot of hand-wringing without action, which ultimately was demotivating for citizen scientists.
Another important consideration was how to make the data collection process more inclusive and whose concerns should be a crucial part of the design. Lakes are public social-ecological systems. Therefore, lake data must not focus narrowly on the concerns of English-speaking residents of gated enclaves. Often, the fisherfolk using the lake were both willing and able to be active participants once they realised that their voices mattered.
Third, people asked questions and we didn't have answers. A big embarrassment for us was this. Once citizens collected data on nitrates and phosphates, when they asked us “is this reading good or bad?”, we weren't able to give them a simple answer. This quickly led to the question, “what’s the point of collecting data if we cannot interpret the results?”. Explanations about water quality depending on nutrient dynamics and designated use were much too complicated to be satisfactory.
We hoped that even if we weren’t able to set a simple quantitative target for nutrient levels in lakes, we would still learn something about seasonality. E.g. we would know how much, if any, dilution was occurring during the monsoon and whether the lakes were able to self-cleanse after a raw sewage “spill”.
In lakes that get raw sewage, the monsoon is a period when the raw sewage gets mixed with stormwater and dilutes sewage enters the lakes. So water quality improves during the monsoon.
However, in other lakes, the municipal corporation — Bruhat Bengaluru Mahanagara Palike (BBMP) — intervenes by adding diversion drains at the lake inlet to prevent sewage from entering the lake. During the dry season, the sewage level remains below the diversion wall and so the sewage gets diverted. However, storm water gets mixed with sewage during the rainy season, causing the levels to rise above the wall and enter the lake.
In these lakes, water quality gets worse during the monsoon. Sewage mixed with storm water enters the lake. But after the monsoon, the lake self-cleanses to some extent as these pollutants settle down and degrade.
While the citizen-collected water quality data provided useful insights, it wasn’t enough to inform action and answer the specific questions citizens were asking us.
Fourth, just because we are water geeks, it doesn’t mean everyone is. Our biggest learning was that people engage with lakes for reasons that go far beyond logical explanations about improving lake water quality. So, engaging citizens in the long-term with such a narrow goal was too limiting. Citizens care about biodiversity, lake quality, access, encroachment, and recreation, and are curious about a wide range of scientific topics. There are also different types of engaged citizens, not all of whom were necessarily interested in science or policy. We needed to engage citizens more holistically and cater to a wider range of interests.
Finally, the process of data collection, curation, and archival must be self-sustaining. This was probably the hardest pill to swallow. Creating public, open datasets that become part of the “knowledge commons” is a valuable endeavour. But in the age of viral posts, where information is consumed in bite-sized posts for short bursts of time, how should more comprehensive scientific data and insights and knowledge of successes and failures be curated so that it is discoverable? We don’t have a clear answer. But the answer is probably that it should be decentralised or be hosted on a common portal like India Water Portal that has longevity and funding.
In Part 3 of this series, we will talk about how we learned that we cannot sustain citizen science centrally and how we transitioned to thinking about lakes as learning spaces — focusing on building capacity to measure, understand and interpret data around lakes.
Stay tuned! You can follow Mira community on Instagram for more on nurturing Bengaluru lakes. Our journey was funded by a CSR grant from Oracle.