In the heart of Dar es Salaam, Tanzania’s six million-strong port city and commercial capital, a dramatic scene unfolded in November 2023 as two municipal workers deftly used a roaring chainsaw to slice up a fallen fig tree at a busy intersection during rush hour traffic.

Dangling in a bucket crane, the duo talked loudly as the blade cut through the wood, sprinkling sawdust into the breeze. This iconic tree had provided shade in the smouldering heat of the city centre and acted as a meeting point for over 75 years. But it was felled one evening during an onslaught of the El Niño rains that have pounded the east African region in recent months.

Dar es Salaam is extremely vulnerable to flooding and heat stress, with recent rains displacing thousands, wrecking infrastructure and exposing the city’s vulnerability to extreme weather. But since 2021, a team of student urban planners have been using cutting-edge artificial intelligence (AI) to devise solutions to bolster the city’s climate resilience.

“We are using AI to create a big map that shows where all of the city’s trees are,” says Lilian Lymo, a student from Ardhi University in Dar es Salaam. Lymo is one of 50 students working on a pioneering tree-mapping project with Resilience Academy, an initiative supported by the World Bank’s Global Facility for Disaster Reduction and Recovery (GFDRR) that engages students from across Tanzania to build capacity in urban planning and climate risk management.

Armed with tablets, smartphones and laptop computers, the five-year project sees the students using open-source labelling tools and high-resolution satellite imagery to map and catalogue leafy assets across the city.

The tree-detection AI then looks at these images and analyses the presence of trees in order to create a detailed map of tree cover density, and this map is currently being used by city authorities to assess the ecological impact of urban development and identify where it needs to plant more trees.

A new era in conservation science

AI tools are being used in conservation to record everything from air quality to tree loss from diseases, but according to Resilience Academy lead researcher Rebeka Mponeja, this particular initiative is harnessing the power of AI to broaden people’s understanding of the city’s ecosystems and to improve the government’s planning and conservation strategies.

“The use of AI for tree mapping transforms data into actionable insights. Through machine learning algorithms, we can predict environmental changes, identify at-risk species, and tailor conservation efforts,” she tells Equal Times.

Trees play a crucial role in reducing air pollution, cooling air temperatures, absorbing carbon dioxide, reducing flood risks and soaking up surface run-off. “When we have a big leafy oasis in the middle of the buildings it will make the city a better place for everyone,” says Lymo.

However, as the population in Dar es Salaam – one of the world’s fastest growing cities – continues to shoot up, rapid urbanisation along with poor planning and deforestation fuelled by widespread charcoal use is taking an increasingly heavy toll on the forests that surround the city, leading to water shortages, loss of biodiversity and increased exposure to disasters.

The Resilience Academy’s tree-mapping project is the first time that green cover has been documented in Tanzania. Not only is Dar es Salaam City Council working closely with the GFDRR to make sure that the data is being used to drive positive action but there are also plans afoot to see the project replicated in other cities across the country, including the capital city, Dodoma.

Mponeja says that AI technology heralds a new era in conservation science enabling data scientists to monitor real-time changes, detect patterns, and swiftly respond to environmental challenges. “AI tools have allowed us to focus on specific tree species and take conservation measures that can make a real difference,” she says.

According to Mponeja, the scheme has set a new standard for urban greening efforts, highlighting the importance of data-driven decision-making and community involvement in building climate-resilient cities. For example, the data amassed by the Resilience Academy has revealed a huge disparity in the tree cover available in different parts of the city, says Mponeja. “We need specific actions to ensure that city dwellers have fair access to green spaces.”

Nick Jones, a GFDRR data scientist working on the Dar es Salaam project, says that the tree-mapping initiative offers a beacon of hope for those who are trying to find solutions to some of the problems posed by urbanisation. “By using data-driven methods to map and keep an eye on the city’s trees, the project creates a strong basis for sustainable growth.” He continues: “By understanding the distribution and health of trees, cities can mitigate the effects of climate change, enhance air quality, and promote biodiversity.”

Mapping to close the inequality gap

One of the outcomes of the tree-mapping project has been the evidence of the yawning gap between tree cover in Dar es Salaam’s affluent neighbourhoods and that of its impoverished townships. According to Jones, while tree canopy blankets over 20 per cent of surface areas in wealthier neighbourhoods, it plummets to less than 0.5 per cent in poorer areas.

“This stark contrast not only highlights socio-economic inequalities but also exposes vulnerable communities to the severe effects of floods and heatwaves,” he says.
With approximately 70 per cent of its inhabitants living in informal and unplanned settlements, Dar es Salaam is highly vulnerable to flooding and intense heat.

In Tandale – a labyrinthine informal settlement of approximately 50,000 inhabitants in the west of the city – local residents say that yearly flooding, of which the lack of tree cover leaves them particularly vulnerable to, has led to the recurring outbreaks of diarrhoea and cholera.

Grace Mwasha, a 35-year-old Tandale resident and mother of four, says that more needs to be done to help people understand the importance of trees and of taking care of their environment, as many low-income households rely on cutting down trees to provide charcoal for cooking fuel. Tanzania loses approximately 469,000 hectares of forest every year due to various factors including tree felling for charcoal, logging and land clearing for farming.

“This place is too crowded. I wish there was more space to plant trees and get fresh air,” she tells Equal Times. “But there is no political will to improve the people’s welfare in this area; that’s why nobody takes the trouble to plant trees here,” she says.

In an interview with Equal Times, Omar Kumbilamoto, the mayor of Dar es Salaam, said the tree mapping initiative will help the city council with future development policies.

“This project shows that the government is determined to make better decisions and build a better future for the people,” he says. “The data collected will serve as a valuable resource to protect the city from the impacts of flooding,” he said, which will help people like Grace.

The mayor also expressed optimism that a data-driven approach would contribute significantly to the city’s overall resilience strategy. “Understanding the distribution and health of trees is important in developing effective policies and initiatives that promote biodiversity, combat climate change, and create a better urban environment,” he added.

The project is ongoing until 2026 but Jones says the success of this initiative so far demonstrates that harnessing the power of data science and technology can make a big difference in a big city and provide a template for other cities to follow.

“The challenges Dar es Salaam faces are not unique, many other cities struggle with deforestation, air quality and heat problems. Implementing similar tree mapping initiatives elsewhere can have a big impact on global urban sustainability efforts,” he
says.