Aadhaar for trees? Meet the people who are building tree maps of Bengaluru

“Just here, we can see eight species of trees — there’s pink tabebuia, rain tree, teak… You can see that bird — parakeet — flying there,” says Vijay Nishanth, pointing out the diversity of life around a street in Jayanagar. “You often see doves, barbets flying around here… With more fruit and flowering trees, you have more birds,” he says.
Nishanth, an urban conservationist is also known as Bengaluru’s ‘Tree Doctor’, believes every tree is valuable. When 17 trees were poisoned for an Apple hoarding in the city earlier this year, he managed to rejuvenate three of them by applying beeswax and orange oil on their trunks. On his smartphone, he shows me before-and-after photos of sidewalks transformed by newly planted trees, and poisoned trees he’s brought back to life.
It’s around 7 am, and we’re waiting for his colleagues from Project Vruksha to show up. Nishanth is the founder of Project Vruksha, a non-profit group tracks every tree and sapling in the neighbourhood.

While the government has spent a great deal of resources on tracking its citizens through Aadhaar — unique identification numbers based on biometric and demographic data of Indian citizens, it hasn’t carried out any official census on trees. What Project Vruksha is attempting is akin to an Aadhaar project for Bengaluru’s trees.
“Our idea is to release a tree terrain of Bengaluru,” says Shariff, technical head and cofounder at Project Vruksha, over a cup of tea and breakfast at a local Udupi restaurant.

Project Vruksha is doing things they way you would imagine they should be done if one had to go about mapping trees in a city in an organised, actionable manner. They have built Vruksha, an online web app that works on a browser, on which they are building a ward-wise tree map of Bengaluru.
On the map, medium and large trees are denoted with an icon of a green tree, saplings are marked as yellow trees, while free spaces where trees can be planted are marked with orange trees. Apart from GPS data, each tree is photographed and ranked on health, and its thickness and height measured and recorded.

Free spaces are marked on the map with recommendations on what size trees can be planted where. “With data, you can decide which trees are strong enough to withstand winds. We have suggested that as well — which places should have big, medium or small trees,” says Nishanth. According to him, it would take a team of 10 people about 18 months to map trees in all the wards of Bengaluru.

“We’re enabling interactivity (on the app) for normal users to register, browse its features; if they want to complain, they can start logging (comments) against a tree, so there will be an activity log for each tree, and on that timeline, we will maintain all its information,” Shariff says.
Also read: Study shows eating red meat, small dietary changes can help India’s water crisis
Their upcoming site revamp is an attempt to try and get everyone engaged with the data. “Let’s assume everyday birdwatching going on around a specific tree, and 10 people are interested in it. They can just collaborate and use the tree as a starting point,” Shariff adds.

An axe to grind

The project, which plans to map all the trees in the city on the site, has been stonewalled by Bruhat Bengaluru Mahanagara Palike (BBMP) officials, and funds from the government are not forthcoming. When we first spoke over the phone, Nishanth shared an article which appeared in the local edition of the Indian Express, in which former deputy mayor S Harish alleges that the project, if funded, would have exposed the prevailing corruption in the BBMP. “We’ve been doing this since 2014. We’re pushing the government (for funds), but it is not agreeing. And you saw the article,” he says.
For now, Vruksha’s tree map is relegated to a handful of wards — Pattabhinagara, Jayanagar east, Byrasandra, and Kanakpura Road — in the city. Nishanth says they believe in quality over quantity when it comes to maintaining data. “If you put up more data, it will become stagnant,” he says. Meanwhile, new features are coming up on the platform in a month, which includes a responsive design for wider hardware support.

A crowdsourced approach

Meanwhile, at the IISc campus in Bengaluru, Gubbi Labs, a research collective, is trying out a crowdsourced approach to carry out a tree census in Bengaluru. Launched as a part of Neralu, a tree festival held in the city in February this year, the tree mapping app is hosted on Github, a repository for code and open data. At present, the map marks trees in Lalbagh, Indiranagar and Cooke Town neighbourhoods, and has sparse data on a few others. Over 7,000 trees have been mapped on the platform.

“Mapping (of trees) has been tried by a lot of people, including the government, which is supposed to do a tree census, but often the challenge is getting people to go around and map (them),” says Dennis C Joy, resident editor at Gubbi Labs. He estimates it would take a minimum of two people for each of Bengaluru’s 198 wards to map and keep track of trees. As this is too resource-intensive, they’re pursuing the other path. “We’re taking the open source way — where you put it out, and kind of hope that people catch on to it, and start contributing (to the mapping exercise),” he says.

Gubbi Labs has an inhouse team of five and a little over a dozen volunteers who are working on the tree mapping project. To supplement their efforts, it is reaching out to schools to help with data collection. Since school kids don’t usually have smartphones, the research group is getting them to use an offline method of mapping preferred by biologists — field papers. “So you have a hard copy (a printed map) of a local area, and you go around marking trees on the paper, and you can digitise the map,” Dennis explains. “A bit ironic — using paper for tracking trees, but you can’t expect kids to have smartphones. You have to have an offline version for them,” he adds.
Since school students are also unlikely to be well-versed with the taxonomy of Bengaluru’s trees, Gubbi Labs organises mapping workshops to help them with identification. Resources, such as S Karthikeyan’s free book on Bengaluru’s flowering trees, are shared in the workshops. While adding a tree, smartphone users can rate their entries on how confident they are about their identification.

“Initially, we were looking at (mapping) only avenue trees, so there are a lot of common trees — with a little bit of learning, you can size them up easily,” says Vignesh Kamath, research fellow at Gubbi Labs.
Also read: Kerala’s first bird atlas: Where citizen meets science, thanks to technology
The collective is trying to combine the crowdsourced data with other methods, such as satellite images, for effective mapping. “We’ve mapped around 3,000 trees like that,” says Dennis. However, overlapping canopy from satellite footage is a major issue in discerning individual trees, says Kamath.
There are many people who want to do something for the environment, so they can start with mapping their part of the neighbourhood,” says Kamath. “It doesn’t take too long either. Once you start doing it, you also get a sense of ownership, a sense of belonging, and you kind of feel the need to protect it,” he adds.

Two roads to take

While Project Vruksha is taking a top-down approach, Gubbi Labs is taking a bottoms-up approach to tree mapping. “There are mainly two branches — one is the conservation side, and then there is us, who come from the science side,” says Dennis.
Gubbi Labs is just mapping trees and is not into conservation, he explains. “Our objective is just to map trees and collect data. We’re not trying to save trees or anything, as of now. It is to build data, so someone else can use it. Once you have data, there are endless things you can do with it,” he says.

However, a detailed look at the map made by Gubbi Labs shows that its mapping project has a long way to go in terms of developing a feature set. The system isn’t geared to track trees that have been cut, and the frontend currently only displays locations of trees — no taxonomical data or photos are shown.
“Right now, what we lack is verification. So, if somebody were to mark a tree by accident, they can’t edit it out,” explains Dennis, adding that having a verification process in place would make their data accurate.

Data to the rescue

Vruksha’s mapping project is making waves in the mapping and open data community, with positive testimonials from the tech community.
“Vruksha is entirely dedicated to tree mapping. Vijay Nishanth is well known as ‘Tree Doctor’. I have full respect for what they do — we’ve tried collaborating with them a couple of times,” says Varun Hemachandran, founder of Talking Earth. The NGO tracks around 20 environmental datasets to find out how it affects people who live in that area. Talking Earth uses three approaches to mapping trees — volunteers, a combination of remote sensing and 360-degree photography, and through private mapping initiatives.

Hemachandran says Gubbi Labs is doing a bunch of mapping projects, and he doesn’t know much in detail about their current work.

He, however, recalls a recent incident in which his NGO collaborated with the collective. “We were called to do a private mapping overnight,” he says. A government veterinary hospital was supposed to come up near Cantonment Station in the city, and the BBMP reported that only four trees were to be axed. Activists found out that the number had been fudged.
“These were only four of 60 trees they were chopping. But these four trees were massive ones — each had a canopy of an acre, and was about a hundred years old,” he says. “That’s when we called Gubbi Labs to help us out,” he adds.
Data saved the day, and the activist team that was objecting to the project took the matter up with the mayor and the BBMP commissioner; the project was halted. Similarly, government estimates of trees to be cut for the proposed steel flyover and the Jayamahal Road widening project were short by more than 500 trees.
Mapping the trees of a city could have many potential uses, and bring a new way to quantify habitability, adds Dennis. “Just looking at habitability from a neighbourhood perspective, they can probably use the app to figure out if a place is green or not. It will help common people as well, and scientists for sure,” he says.
Also read: Young ecologists use machine learning to prove that more birds are endangered in the Western Ghats than we know