TreesAI is implementing location-based scoring in Stuttgart
Trees-as-Infrastructure (TreesAI) was evolved as an initiative by Dark Matter Labs with contributions by Lucidminds and is now being held by Dark Matter Labs. TreesAI is exploring the required organisational infrastructures to revalue nature as a critical part of urban infrastructure alongside bridges, roads and rail. TreesAI is part of a wider Nature-based solution mission at Dark Matter Labs, focused on supporting nature-inspired approaches that sustain and regenerate the health of underlying ecosystems.
In this blog we summarise how one of our tools, the TreesAI Location-based Scoring, has been applied in Stuttgart to assess climate risks spatially and support the design and prioritisation of different urban nature-based solutions. None of this would have been possible without the valuable collaboration of various partners in Stuttgart: Bernd Junge, Clemens Hartmann, Ekkehard Schäfer, Elisabeth Bender, Fridtjof Harwardt, Hauke Diederich, Holger Wemmer, Jan Kohlmeyer, Johannes Wolff, Juliane Rausch, Katja Siegmann, Niels Barth, Sophie Mok, Sven Baumstark.
TreesAI Stuttgart is cofunded by Stuttgarter Klima-Innovationsfonds and The Nature Conservancy
(1) Introduction
Since 2023, TreesAI has been working in Stuttgart, Germany, after successfully applying to the nature-based solutions funding line by the Stuttgarter Klima-Innovationsfonds and The Nature Conservancy.
TreesAI is being implemented in partnership with several city departments and municipal facilities: the Stuttgart climate protection unit, the urban planning department, the civic engineering office, the health department, the tree management teams for city trees, and trees on state premises.
The project kicked off in September 2023 with an in-person, cross-divisional workshop with all involved city departments in Stuttgart to align on project goals.
Two key goals came out of the workshop. The first one was around the quantification of benefits trees provide — for which Stuttgart is using GUS which is essentially a comprehensive, AI-powered platform created to support stakeholders in forestation projects. It accommodates various forest types, scales, and geographic locations, maintaining scientific rigour with its peer-reviewed framework. GUS is now being developed and maintained by Lucidminds.
The second goal, which is this blog's focus, was to synthesise existing data to support decision-making around the location and prioritisation of projects. In this blog, we share our experience to date of working with the city departments through a co-created process of scoring projects through their location — what we call “Location-Based Scoring”
TreesAI Location-based Scoring (LBS) provides an overlay and weighting process of location-based vulnerabilities to climate risks in the city — like heavy rainfall hazards, air pollution and heat islands — and helps decision-makers assess the most effective locations for maintaining and increasing green infrastructure to mitigate climate risk across the city.
How can Location-Based Scoring (LBS) be used to make the case for urban nature?
The LBS methodology applied in Stuttgart is a tool designed by TreesAI to perform a risk-based vulnerability assessment. This helps to evaluate how patterns of risks and potential benefits on natural and human systems are shifting due to climate change.
Climate risks result from an interplay between hazards, stress factors, exposure and vulnerability. Vulnerability is special in that it is not only determined by the sensitivity to damage yet also by the coping capacity to deal with it.
The overall aim of this analysis was to facilitate an understanding of:
- Hazard-Exposure relationships of climate change in the local context.
- Identify geographical hotspots that lack green in relation to their level of risk and potential.
- Consolidate compound site factors — including population density, surface water flooding risk, heat island effect, air pollution concentration, site accessibility — into a single overall location-based score.
For this purpose, the LBS methodology involves developing a risk score for geographical areas based on the different focus areas in Stuttgart, and it is a methodology adapted from the “Impact and Vulnerability Analysis of Vital Infrastructures and built-up Areas Guideline” (IVAVIA) (Resin, 2018) that is based on the concepts of risks defined by the IPCC’s Fifth Assessment Report.
LBS reveals which areas would highly benefit from specific Nature-based Solution typologies to increase their coping capacity to climate risks in the city.
(2) The story so far
Scoping assessment in Stuttgart: defining focus climate themes
The first step of the onboarding process for Stuttgart was the definition of the most pressing risks in the city’s local context that NbS could help mitigate and adapt, which could also damage green infrastructure. In other words, this step was about defining the hazards and exposed assets or systems for assessing risks in the city of Stuttgart. Therefore, the TreesAI team, in partnership with the involved city departments and municipal facilities, jointly crystalised the following five hazard-exposure relationships:
- Heat Risk on Population Health
- Drought Risk on Green Infrastructure
- Air Pollution Risk on Population Health
- Surface Water Flooding Risk on Built-up Areas
- Surface Water Flooding Risk on Transport Network
For each hazard-exposure combination, an impact chain was crafted. These impact chains describe the cause-and-effect dynamics between hazard and exposure components. Subsequently, indicators were established to delineate the three primary facets for assessing climate risk: hazard, exposure, and vulnerability, which includes coping mechanisms and sensitivity. Each of the five impact chains and the selection of these indicators were guided by the data available in Stuttgart and, crucially, by co-creation workshops and interviews with experts, policymakers and through scientific research.
Given that LBS focuses on spatial analysis to rank NbS according to their necessity gathering spatial data on pertinent urban indicators was an essential stage of this partnership with the city. The team secured a variety of inventory data from the city alongisde open-source satellite data, encompassing aspects like population density, surface water flooding risk, urban heat island effect, the extent of impermeable surfaces in built-up areas, and a tree register. These data sets were then synthesized using the TreesAI LBS methodical approach.
An impact chain diagram shows the relationship between these components. Below is an example of the impact chain diagram developed for Stuttgart to visualise the factors needed to evaluate the impact of heat (hazard) on population health (exposure)
Scoring the city according to NbS potential to mitigate risks
After gathering all the necessary data and establishing the analysis resolution, initially set as a 500x500m grid, we processed all data to fit this spatial framework using GIS software. This processing allowed us to compute a location score for each spatial unit by calculating the various indicators of risk components in the LBS Modeling. This calculation involved weighting, normalizing, and aggregating the different indicators.
The weighting of indicators enabled us to prioritize among risk components, such as hazards, vulnerabilities, and exposures. In the case of Stuttgart, the assignment of weights to these indicators was achieved through collaborative workshops with the city’s municipal departments. Each workshop focused on a specific impact chain, gathering representatives from departments relevant to the climate theme under discussion. For example, in the workshop addressing the risk of drought to green infrastructure, the attendees included experts and stakeholders directly involved in this area, including the Stuttgart tree management teams for city trees as well as trees on state premises, the Stuttgart climate protection unit and the Stuttgart urban planning department.
During these sessions, we presented all the indicators, and participants allocated points based on their perception of the indicators’ importance and relevance to reflect the level of risk.
User Interface
To present the results of the LBS in spatial data, maps are produced using a Geographical Information System (GIS) and a web-based interface is developed to provide an interactive dashboard geared to the needs of the city of Stuttgart.
Maps can effectively present geographical comparisons of climate damage in the city for either spatial unit. Charts can also illustrate the combined risks of one hazard or show the risks of one impact chain in the city.
The platform features two approaches for the spatial analysis (a) “Explore the risk of a location” and (b) “Find locations (with characteristic X)”.
The dashboard function “Explore the risk of a location” analyses each (project) location in the Stuttgart city area preferred by users, supported by an address field input and zoom function.
(3) What comes next
Now that we have the platform for Stuttgart, there are various opportunities for the municipality to choose to use the results to support decision-making and enable more effective development, maintenance, and monitoring of NbS.
- Guide the location of NbS by identifying areas susceptible to climate risks. LBS can potentially connect different urban structures with specific Nature-Based Solution (NBS) types, optimizing resilience strategies tailored to distinct risk profiles.
- Guide the design of NbS by pointing out the location's risk profile, for example, choosing SuDS-enabled trees in areas of flood risk.
- Guide the development of maintenance schedules, as LBS can highlight areas where existing trees are at higher risk of drought or pests due to factors like soil conditions, urban heat islands, or insufficient care.
- Guide policy by providing data to policymakers for creating and implementing environmental regulations based on spatial risks.
Beyond this, we are exploring how LBS could be linked to ecosystem valuation tools to support a more radical and collaborative approach to delivering and financing urban forestry, such as:
- Raising private capital as outcome payments for the benefits provided by trees.
- Supporting private land-owners to contribute to risk mitigation through nature-based adaptation.
- Assist utility companies and infrastructure developers in choosing optimal locations for integrating green infrastructure in their locations.
- Integrate LBS into real estate databases or insurance risk assessment models
- Repurposing budgets from preventive health initiatives to fund NBS
If you are based in Stuttgart and want to get involved, or if you are from another city and want to find out more, don’t hesitate to get in touch at: treesai@darkmatterlabs.org
TreesAI LBS in Stuttgart: Sebastian Klemm, Chloe Treger, Sofia Valentini, Gurden Batra
GUS in Stuttgart: Oguzhan Yayla, Bulent Ozel, Cynthia Mergel and Jake Doran
Platform Design & Code: Arianna Smaron, Alessandra Puricelli, Gurden Batra
