SPADE

The effects experienced by climate change are increasing in frequency, causing the occurrence of extreme weather. One example is heavy precipitation, or single-day precipitation events, which present material and life risks through large flows of water and rapid flooding.

According to EPA (2021), since the 1980s, the prevalence of extreme single-day precipitation events has risen substantially, with nine of the top 10 years for extreme one-day precipitation events occurring since 1996. Between 1910 and 2020, the portion of the country experiencing extreme single-day precipitation events increased at a rate of about half a percentage point per decade.

A study from Davenport et al. (2021) revealed that there is an increase in precipitation levels and historical changes in precipitation contributed to about one-third of flood damages from 1988 to 2017, which represents an additional impact of 73 billion dollars in damage during this period.

As explained by the United States Geological Survey, intensity and duration of rains, and topography of the precipitation area are among the variables used to predict floods. While the topography of the region is a fundamental element in describing the path taken by rainwater (and its accumulation), the ground cover is a determining factor in understanding what portion of the water will be absorbed and how much will flow to lower regions.

Although bibliography presents some studies for predicting the behavior of rainwater and its possible implications, these are often punctual (they deal only with a certain area, usually large urban centers), difficult to access and understand for the population general or, in many cases, outdated due to the great speed in which the changes in the landscape occur.

Nowadays, there are mechanisms, many of them open sources, that provide detailed and updated information about the topography and land cover of different parts of the world (such as ESA's Copernicus project and NASA's SRTM and ASTER). These data can be applied to update rainwater flow paths and flood maps from previous studies (accounting for recent changes in topography and land cover) or generate new maps of regions that, due to the high costs of previously existing techniques, were not studied.

However, despite the public availability of this data, the analysis and study of heavy precipitation impacts are still specific and target certain audiences. Through the large-scale application of these data (generating simple and easy-to-understand results), governments, companies, and individuals could not only apply it for disaster management but also as a decision support system for everyday decisions like drainage design, insurance values, or buying a property.

Sources:

EPA (2021)

Davenport et al. (2021)

United States Geological Survey

Our solution is called Simulator for Predicting Accumulation and Drainage through Elevation (SPADE). It is an easy-to-use tool that provides insights about rainwater flow paths and areas susceptible to floods.

SPADE applies open-data Digital Terrain Model (DEM), satellite images, and user's input regarding the precipitation and desired study area location to provide a representation of rainwater behavior on an interactive map.

Initially, our solution applies DEM to create a point cloud of city elevations. The cloud is later used to define local water basins and the elevation neighborhood of each point. This data serves as input for an algorithm that calculates probable rainwater flow based on the difference in elevation of the cloud points. The algorithm also determines points with a lower elevation within a group of neighbor points, indicating a potential water accumulation area and, therefore, a region at risk of flood.

The system also applies image recognition techniques to analyze satellite images to classify the area's different types of surface cover. Each type of coverage has an associated permeability index. This index is important to calculate how much of the water that falls on the land is absorbed and how much is transferred to areas of lower elevation. The application of this calculation makes it possible to understand the volume of water that flows through a certain region, as well as the height that the water surface can reach at points at risk of flooding. It also makes it possible to study the dimensioning and positioning of drainage mechanisms to analyze their impact in situations of heavy precipitation.

SPADE uses open data, such as those provided by the GeoE3 platform, ESA's Copernicus program, and the CGIARCSI consortium. Our algorithm works with no pre-processing, calculating the information at the moment it is requested by the user. This means that it can work with the most recent data available. This characteristic is vital for providing an assertive analysis of the ever-changing urban landscape.

In summary, SPADE is a tool that provides insights into the effects of heavy precipitation in an area. By using location-based data and advanced algorithms, SPADE can accurately predict water flow paths and areas susceptible to floods, and provide real-time information to stakeholders. This solution can assist in minimizing the impact of flooding on people and communities in urban areas, as well as predict the effects of human activities on rainwater flow and accumulation.

A prototype of SPADE is already online at: SPADE prototype
The functionality of the system can also be checked in: Video of SPADE in action

SPADE is easy to use. The user inputs the desired city and the prediction of the amount of rain and presses the process button. After a short wait, SPADE provides the probable water flow, possible accumulation points, and the volume of water accumulated. Right now, our prototype can study the areas of Coimbra and Porto (Portugal) and apply simplified satellite images for determining the surface cover, like the ones from the OpenStreetMap app.

Our solution can attend to the needs of a variety of users. Since it is designed to be easy to use, its basic features are available to any person or company. Risk analysis related to heavy rainfall can be applied in various areas.

Rescue services can foresee which areas will be in danger in the event of heavy precipitation and plan ahead. Governments can use SPADE to design drainage systems, certify areas for construction and study the impact of landscape or soil cover changes. Insurance companies can use SPADE to study the area surrounding a building and better assess its risks. Alternatively, people who are seeking for buying/renting a place to live or work can use our tool to understand how the region behaves in extreme weather events.

Therefore, SPADE is a powerful tool to assist communities understand and incorporate extreme climate events, and their risks, in infrastructure design and planning.

The impacts resulting from heavy rainfall can be felt by the entire population, which includes our team, who have lived and witnessed these problems in cities such as Florianópolis, Fortaleza, São José, Coimbra, Porto, and Calgary. The SPADE team is composed of two doctoral students with a background in civil engineering and experience in optimization, artificial intelligence, and, especially, scientific research. Applying our background in addition to personal experiences, we were able to perceive how these problems may impact the overall population through a bibliographic review of academic papers and governmental reports. This analysis was extremely important to direct the development of SPADE. This way, we could collect the knowledge produced by other researchers and add their perceptions to our own.

At this point, SPADE had the validity of its idea attested by a second place in the 2023 Location Intelligence for Smart Cities Hackathon. In the Hackathon, SPADE was praised for its versatility, being able to be applied in a great number of applications, all of them of great importance. The jury for this event was made up of specialists in urban planning, geoprocessing, and government representatives. The award of second place was seen by us as an indication that we are in the right direction to help solve a problem that affects the social, economic, and political aspects of the population.

Furthermore, throughout the maturity and development of the prototype, we intend to work with governments, insurers, emergency services entities, and the population in general. This close contact will serve to evaluate new possible applications of SPADE and implement a wider range of tools.

The basic coding for predicting rainwater flow, water accumulation, and soil cover identification is ready. Those are integrated and functional in a platform available at:

Anyone can access our platform and test our algorithm. Right now, their calculations are available for the Portuguese cities of Porto and Coimbra, but SPADE's cover area can be easily expanded.
Moreover, we consider that our ideas, methods, and platform were validated after we got second place in the 2023 Location Intelligence for Smart Cities Hackathon.

At this moment, our platform covers the Portuguese cities of Porto and Coimbra, which combined have a population of about 600.000 inhabitants. Once our prototype is free to use, it can be used not only by the inhabitants of these two cities but by everyone who wants to study the effect of heavy rains in those two cities.
Our code was designed to be dynamic and new cities can be easily included. Therefore, the population covered by SPADE can be quickly expanded.

We are applying to MIT Solve because being part of the Solve community can provide us with valuable resources, partnerships, and support to help us overcome our current barriers. Our project aims to prepare communities and cities for the changes and extreme weather events that will occur in the coming years, and we need partners interested in expanding our solution worldwide.

While we can expand in Europe due to the availability of open DTM data, we are looking to reach out to other countries and regions interested in using our technology, such as insurance companies, cities at risk, and areas with potential flood risks. Climate change-induced flooding is a real and growing phenomenon that increasingly affects urban regions worldwide.

We believe that Solve can connect us with potential partners, grants, or investors who can provide financial support to develop our solution further and expand its reach, especially outside the EU, where we would like to offer our services in regions with available data. As we work with open data, our financial costs mainly involve maintaining cloud servers for storing and computing data, as well as the costs associated with developing and integrating new databases into the solution.

In addition to financial support, we face challenges with legal issues, establishing ourselves as a company or organization, and navigating the complexities of growth. Solve can provide guidance and resources to help us overcome these legal hurdles and ensure a strong foundation for our solution as it grows. By becoming a Solver team, we want to leverage the diverse resources, expertise, and network that the MIT Solve community offers to overcome these challenges, scale our solution, and make a lasting impact in the fight against climate-induced flooding.

In simple terms, SPADE is capable of analyzing and transforming complex open-source data into a simple and interactive format for the most diverse applications.

One of the main aspects that make SPADE innovative is to present a simple and dynamic interface to the user, making it accessible to different types of users (governments, companies, and ordinary people), allowing access and dissemination of knowledge about the area they inhabit, as well as its application in diverse areas and purposes.

Another point that should be highlighted is the constant updating of the information presented. The speed with which humanity modifies the environment can disassociate older studies from reality. SPADE is capable of carrying out new studies in an easy and fast way, from the supply of new updated input data.

For the next year, our goals to SPADE are:
- Expand the coverage area and user base of SPADE: We aim to expand the coverage area of SPADE to make it accessible for larger groups. Our algorithm was developed to be easy to expand, so the technical part should not be an issue and this should be a natural course of the project, given time and support.
- Improve the user experience and develop new features and functionality: SPADE is useful in various scenarios. However, at this point, it is a prototype. User experience can be improved by improving the design of the platform and implementing other functionalities, such as the impact of drainage points and water level calculation.
- Enhance the accuracy of predictions: consulting coding and engineering specialists, we aim to upgrade our code to increase the accuracy of our predictions.

In the long run, we seek to:
- Partner with organizations: We intend to work with governments, rescue services, ONGs, and all sorts of organizations to adjust SPADE to make it into an increasingly capable tool.
- Make SPADE a known tool and promote its use, through marketing and partnerships.

Some UN's SDG that relates directly to SPADE are:

• 11.5 By 2030, significantly reduce the number of deaths and the number of people affected and substantially decrease the direct economic losses relative to the global gross domestic product caused by disasters, including water-related disasters, with a focus on protecting the poor and people in vulnerable situations
• 13.1 Strengthen resilience and adaptive capacity to climate-related hazards and natural disasters in all countries
• 13.2 Integrate climate change measures into national policies, strategies, and planning.

The correct way to measure the impact of our system is to compare historical data regarding heavy precipitation impacts with new ones in the areas where SPADE is available and used. Some indicators that may be used:

• Number of flooded areas during heavy precipitation;
• Average time for rescue services to reach areas in need;
• The yearly value of material losses related to heavy precipitation events;
• The yearly number of life losses related to heavy precipitation events.

Our theory of change is centered around the belief that by providing accurate, timely, and accessible flood predictions, we can empower communities to better prepare for and respond to the challenges posed by extreme weather events, ultimately mitigating the adverse effects on lives and property.

To achieve this impact, we have designed our solution, SPADE, which combines Digital Elevation Models (DEMs) and satellite imagery analysis to predict rainwater flow, identify flood-prone areas, and assess rainwater infiltration. By continuously refining our model through advanced GIS technologies and artificial intelligence, we aim to improve its predictive capabilities and accuracy.

The immediate outputs of our solution include:

1. User-friendly, interactive maps, and visualizations help users understand the flood risks in their area.
2. Information that enables governments, rescue services, insurance companies, and the general public to make informed decisions regarding flood preparedness and response.

Our longer-term outcomes include:

1. Improved community preparedness and response to extreme weather events, leading to a reduction in loss of lives and property damages.
2. Improved urban planning and infrastructure development by integrating flood prediction insights, resulting in more resilient communities.
3. Greater public awareness and understanding of climate change-induced risks, fostering a proactive approach to disaster management and environmental sustainability.

The logical framework that connects our activities to the immediate outputs and longer-term outcomes is as follows:

1. Activities: Develop and refine SPADE using advanced GIS technologies, AI, and satellite imagery analysis.
2. Outputs: Accurate flood predictions and easy-to-understand visualizations, leading to better decision-making by various stakeholders.
3. Outcomes: Enhanced community preparedness, improved urban planning, and increased public awareness about climate change-induced risks.

Evidence supporting our theory of change includes third-party research on the effectiveness of flood-predicting tools and data analysis from pilot SPADE implementations in selected cities. Furthermore, we intend to consult stakeholders soon. Through continuous feedback loops and iterative improvements, we aim to strengthen the links between our activities, products, and results, ensuring maximum impact on the issue.

The core of our solution is powered by a combination of Digital Elevation Models (DEMs) and satellite imagery analysis. We leverage advanced GIS technologies and artificial intelligence to create a predictive tool, named SPADE (Simulator for Predicting Accumulation and Drainage through Elevation).

SPADE analyzes elevation data and satellite images to calculate probable rainwater flow, pinpoint areas prone to flooding, and assess the extent of rainwater infiltration. In the next phase of our development, we plan to incorporate Artificial Intelligence for satellite image segmentation. Currently, we use computer vision models and process images on a pixel-by-pixel basis. By integrating AI into our solution, we aim to further enhance the accuracy of our terrain analysis.

Our system, built using Python 3.10.9, operates in the cloud and can easily adapt to platforms like AWS, Azure, or Google Cloud. This flexibility enhances scalability and reliability, ensuring optimal performance for users worldwide.

The impacts of heavy precipitation, rainwater flow, and floods can be seen in all communities, affecting especially vulnerable populations.

Initially, we thought of applying it to assist drainage projects in governments and route planning for rescue teams in an emergency. Its use for the general public (for analysis of places to live), as well as for insurance companies, was later proposed by a professional from another country who, due to his experience and knowledge different from ours, was able to see new applications in which our platform could be used.

Currently, SPADE's team is small, formed by two people who, despite adding knowledge and experiences from different lives, are still close (sharing countries of birth, residence, and academic backgrounds). However, with the growth of our team, we intend to make it as diverse as possible, so that the life experience, knowledge, and views of these people can propose new uses and applications of the system, directing our development towards something that can be useful for everyone. all people.

Since SPADE is a virtual platform, and our team is used to working remotely, we intend to diversify our team with professionals from different countries, forming a multidisciplinary and diverse team.

Our business model is centered around providing value to our users by helping them assess and mitigate the risks associated with climate-induced flooding. Our primary customers and beneficiaries include local governments, urban planners, insurance companies, and citizens living in flood-prone areas.

Our flood risk assessment platform is built on cloud-based servers and uses open-source DTM data, satellite images, and machine learning algorithms to provide information about potential flood risks. For governments and urban planners, we offer information that can be used to increase the resilience of their communities. For insurance companies, we provide a premium API service that can be integrated into their existing systems to help them better understand and manage the risks associated with flooding. For citizens and communities, we provide information that can serve to predict and prepare. Additionally, for rescue groups, we can help with rescues by showing the best rescue routes and risk zones.

We will strive to keep the system free for communities and individual users by implementing request limits. This approach ensures accessibility for those who need it most while managing our processing costs in the cloud. By partnering with international organizations, governments, NGOs, and philanthropic institutions, we aim to secure funding and grants to subsidize the cost of our services for these users, maintaining the financial sustainability of our business.

We will offer API access with costs per request for premium users, such as insurance companies and large corporations. This would allow their automated systems to seamlessly integrate with our platform and access real-time flood risk data, providing them with the information needed to make informed decisions and better serve their clients. The revenue generated from these premium services would help cross-subsidize the cost of our services for low-income users and at-risk communities.

As we continue to develop our solution and expand its reach, we recognize the importance of working closely with local communities and leveraging their knowledge and expertise. To that end, we are exploring options for incorporating crowdsourced data into our platform, allowing users to contribute valuable local information and further refining our risk assessments and adaptation recommendations.

In summary, our business model is focused on providing valuable flood risk assessment and adaptation services to a wide range of customers and beneficiaries, from local governments and urban planners to insurance companies and citizens living in flood-prone areas. By offering tiered pricing structures, implementing request limits, and partnering with organizations that share our mission, we aim to make our services accessible to those who need them most while maintaining the financial sustainability of our business. By integrating advanced technology, local knowledge, and strategic partnerships, we hope to make a lasting impact in the fight against climate change-induced flooding and contribute to a more equitable and resilient future for all.

Our plan for achieving financial sustainability involves a combination of revenue streams that would enable us to cover our expenses while expanding and improving our services. Our primary revenue sources include offering premium services to businesses, securing service contracts with governments and institutions, and leveraging partnerships for donations and grants.

Premium services for businesses: Insurance companies and large corporations would benefit from our flood risk assessment platform, which can help them better understand and manage the risks associated with flooding. We will offer a premium API service with costs per request for these clients, allowing their automated systems to integrate with our platform and access real-time flood risk data. We also plan to provide more accurate results for premium users, offering data for meshes every 5 meters and an option to upgrade to a super-resolution of 1 by 1 meter. The revenue generated from these premium services will contribute significantly to our financial sustainability.

Service contracts with governments and institutions: Local governments and urban planners are among our key customers, as they require reliable flood risk data and adaptation recommendations to build resilient communities. We will actively seek service contracts with governments and institutions to provide our services, generating a stable source of income that can cover our operational expenses. For government entities, we intend to offer the ability to carry out simulations with the addition of drainage systems, enhancing the value of our service for their specific needs.

Donations and grants: As our solution aims to benefit at-risk communities and low-income citizens, we will actively pursue partnerships with international organizations, NGOs, and philanthropic institutions to secure funding and grants. These funds will help us subsidize the cost of our services for users who cannot afford them, ensuring that our solution remains accessible to those who need it most. Additionally, these partnerships can help us access valuable resources and expertise that can aid in the development and scaling of our solution.

Community access: We are committed to offering our solution for free to the community whenever possible. We will implement a request limit for free users to manage processing costs in the cloud. This approach allows us to balance the accessibility of our service with the financial sustainability of our operations.

Our financial sustainability plan primarily focuses on these top three revenue sources: premium services for businesses, service contracts with governments and institutions, and donations and grants, while also ensuring free access for the community. By leveraging these revenue streams, we aim to cover our expected expenses and ensure the long-term viability of our solution. This approach will allow us to concentrate on achieving our mission of helping communities and cities prepare for the challenges and extreme weather events that lie ahead.

Our project has shown promising results so far, receiving recognition and proving the quality of our approach, which in turn indicates the potential of the solution. Winning second place in the 2023 Location Intelligence for Smart Cities Hackathon, we earned a prize of 2,000 euros and an invitation to present our solution at dataweek.de in Leipzig on June 28, 2023. This achievement provided initial funds to support our work and helped raise awareness and connect us with potential partners and collaborators.

Our solution is currently in the prototype phase, so we have not sought extensive funding. However, our solution is designed to be low-cost and scalable, allowing us to make progress with limited financial support. The main factor we need now is to test the solution and gauge users' acceptance.

One of our main objectives is to bring our solution to market and work closely with partners and users to better understand the problems they face and tailor our product to their demands. Our idea holds great promise, but we must validate its effectiveness by engaging directly with users and evaluating their experiences and feedback.

We plan to build on our early successes and leverage the resources, experience, and network available through the MIT Solve community to achieve financial sustainability. Our next steps include creating the enterprise-grade API service. Also, we are seeking new service contracts with governments and institutions that share our vision. By engaging with users and adapting our solution to meet their needs, we are confident that we can further solidify our financial sustainability and ensure the long-term viability of our solution.