Mapping Disasters in the Face of Climate Change: São João del-Rei

This blog is available to read in Spanish and Portuguese.

This blog post aims to present the procedures and results of the project "Mapping Disasters in the Face of Climate Change: São João del-Rei." The study area is the Colônia do Marçal Census Sector, located in the city of São João del-Rei, State of Minas Gerais, Brazil. The Sector comprises 15 neighborhoods and has grown by more than 320% in the last decade, expanding into areas with high and low altitudes and slopes, such as mountains and riverbanks, respectively. Its stormwater and wastewater collection network has been expanded but not modernized, resulting in the main connection networks being overloaded with water flow, lacking maintenance to accommodate a greater volume than originally planned during the implementation project (Zacharias, et al. 2021). The lack of maintenance, increased impermeability of areas, and river canalization result in street flooding during each rainy season, hindering vehicle and pedestrian traffic and putting the population's lives at risk. Figure 1 shows the location of the study area.

The general objective of the project was to map the streets susceptible to flooding and inundations in the Colônia do Marçal Sector, enabling the issuance of alerts during the rainy season. The specific objectives were to collect data using Street Complete, Mapillary, and Kobo Toolbox; gather information from local newspapers and social media about the flooding and inundations in the Sector; and indicate the risk levels of each street during the rainy season in the Sector.

The project also aims to contribute to Goal 11 of the 2030 Agenda, which is to make cities and communities more inclusive, safe, resilient, and sustainable. The goal, by 2030, is to reduce the number of deaths and economic, social, and environmental losses (UN, 2015). In Brazil, this is an immense challenge, especially in the face of climate change, which has intensified the occurrence of floods and inundations, increasing the losses and deaths of people and animals with each rainy season.

Over the years, the rapid urban growth of the Sector into areas with low and high slopes and the inefficiency of the infrastructure for stormwater and wastewater collection have resulted in flooding on the main streets of the Sector, such as Avenida 31 de Março, Avenida Luiz Giarola, Avenida do Contorno, Rua Domênico Randi, Rua José Eugênio de Almeida, Américo Longati, and Rua Monsenhor Silvestre de Castro.

In this context, the project utilized free Geographic Information Technologies, such as Mapillary, OpenStreetMap, and QGIS, as well as a Remotely Piloted Aircraft, to map the streets and avenues susceptible to flooding and inundations in the Colônia do Marçal Sector and make them available in an online repository (GitHub).

2. Data Processing

The methodological procedures comprise four main stages according to the proposed timeline:

1) Data collection using Mapillary;

2) Data gathering from local newspapers;

3) Aerial survey of the area using an RPA (Remotely Piloted Aircraft);

and 4) Creation of a GitHub repository.

Stage 1: Data Collection with Mapillary Data collection using the Mapillary app was carefully planned to ensure the quality of the information. Images were captured using smartphones, covering streets and avenues previously identified as susceptible to flooding and inundations. Initially, the best times and days with less traffic were determined for data collection. Given the extensive length of the mapped streets, the routes were covered using vehicles (cars). The mapping was carried out on the following roads: Avenida 31 de Março, Avenida Luiz Giarola, Avenida do Contorno, Rua Domênico Randi, Rua José Eugênio de Almeida, Américo Longati, and Rua Monsenhor Silvestre de Castro.

Stage 2: Data Gathering from Local Newspapers Historical data on flooding were researched on websites and social media platforms (Facebook) that provide local information, such as Jornal PopNews, Rádio Emboabas, Rádio Campo das Vertentes, YouTube videos, and posts from residents on Facebook, as well as consulting databases from previously conducted projects in the study area.

Stage 3: Aerial Survey Using RPA To generate a quality and up-to-date product, since the satellite image available on OSM seems to date back to before 2019, aerial photographs were collected using a DJI Mini 2 Remotely Piloted Aircraft (RPA). The flight planning was done using QGroundControl software and the Litchi app, available for Android and iOS systems. During the data collection, a spatial resolution of 6 cm was adopted, with longitudinal and lateral overlaps of 75% and 65%, respectively. The established parameters were: speed of 8 km/h, flight altitude of 85 meters above ground level, maximum flight time of 25 minutes, image capture interval of 14 seconds, and camera positioning at 90º relative to the ground (Oliveira, Ventorini, 2023).

The aerial photographs were processed using OpenDroneMap (ODM) software. OpenDroneMap is an open-source software designed to process aerial images captured by RPAs and convert them into geospatial products such as orthomosaics, 3D models, and point clouds. Processing in OpenDroneMap involves several stages, including:

• Importation: The images captured by the RPA are imported into the software.

• Correction: ODM calibrates the images based on flight parameters and the optical characteristics of the camera, correcting distortions and adjusting image geometry.

• Feature Detection and Matching: The software detects and matches common points of interest between images to calculate the position and orientation of the photos.

• 3D Reconstruction: Using the matches between the images, ODM creates a dense point cloud and a 3D model of the area.

• Orthomosaic Generation: The software uses the point cloud and image data to generate high-resolution orthomosaics, which are orthorectified images (i.e., corrected to eliminate distortions).

Once created, the orthophotos are prepared for publication on Open Aerial Map. First, the orthophotos are converted to a format compatible with the platform, usually TIFF or JPEG, and organized into appropriate geospatial layers. It is essential to ensure that the metadata associated with the images, such as capture date, geographic coordinates, and information about image quality, are included and correctly formatted.

Next, the orthophoto is uploaded to the Open Aerial Map platform. The upload process includes entering the metadata and configuring attributes that will facilitate the viewing and querying of the images by users.

3. Results and Conclusion

The data collection using the Mapillary app covered an area of 18 km and resulted in a total of over 3,500 photographs. During the capture process, challenges were encountered due to the poor condition of the streets, with many potholes. These conditions caused jolts to the vehicle, which altered the camera's position, making continuous image capture difficult on Avenida 31 de Março, Avenida Luiz Giarola, Avenida do Contorno, Rua Domênico Randi, Rua José Eugênio de Almeida, Américo Longati, and Rua Monsenhor Silvestre de Castro, located in Colônia do Marçal and Colônia Giarola (Figure 2).

Local data sources, such as newspapers and residents' posts, were essential for validating and complementing the mapped information. The historical analysis indicated that the areas most affected by flooding and inundations in the Colônia do Marçal Sector are: Avenida Luiz Giarola, Avenida do Contorno, and Avenida 31 de Março. These critical points are located near streams that have been canalized and have a high degree of impermeabilization. As a result, all the drainage from the

region is directed to these streams, overloading the system and leading to frequent flooding (Figure 3).

In addition to these factors, intense urbanization without proper planning for

stormwater and wastewater infrastructure exacerbates flooding. The

impermeability of the soil prevents rainwater from being naturally absorbed,

increasing the volume of surface runoff that is directed into the stream. This situation is worsened by the lack of regular maintenance and cleaning of stormwater galleries, which are often clogged with trash and debris, further reducing their drainage capacity. Figure 4 shows an example of a canalized stream in the study area.

The resulting orthomosaics were made available on Open Aerial Map (Figures 5 and 6), which is an open-source platform that allows for the publication, sharing, and visualization of aerial data, such as satellite images and orthomosaics. After being uploaded to OAM, the orthomosaics become accessible for consultation and analysis through an interactive map interface, making it easier to access and integrate the data with other Geographic Information Systems (GIS).

The combined use of different data collection methods, such as Mapillary, historical data gathering, and the use of Remotely Piloted Aircraft (RPA), was effective in mapping and analyzing areas susceptible to flooding, providing a comprehensive geospatial view of the studied area. The aerial images processed in OpenDroneMap allowed for the creation of high-resolution orthomosaics. The integration of these various data sources enhanced the accuracy of the obtained information.

4. Final Considerations

The project "Mapping Disasters in the Face of Climate Change: São João del-Rei" aimed to present a detailed analysis of the procedures and results obtained in the study area of the Colônia do Marçal Census Sector in São João del-Rei, Minas Gerais. Through the mapping of streets and avenues, the main roads susceptible toflooding were identified.

The results show that the residents of the Colônia do Marçal Sector face significant risks during the rainy season, especially related to flooding. The analysis of historical data, together with geospatial information, allowed for the indication of the risk levels of streets and avenues. It is concluded that the project's objectives were achieved and that the mapping can contribute to the generation of alerts during periods of intense rainfall.

Authors

Júlia Gabriela Fernandes Teixeira - “Discente em Geografia na UFSJ e líder do capítulo Unificar Ações e Informações Geoespaciais (UaiGeo) da YouthMappers. Como bolsista PIBEX, participei do projeto "Mapeamento Colaborativo do Patrimônio Arquitetônico de São João del-Rei" em parceria com o IPHAN. 2024 YouthMappers Leadership Fellow.’’

Profa. Dra. Silvia Elena Ventorini - “Possui graduação Licenciatura em Geografia (2004), mestrado em Geografia (2007) e doutorado em Geografia (2012), pela Universidade Estadual Paulista Júlio de Mesquita Filho (UNESP) - Campus de Rio Claro. Atualmente é Professora Associada II da Universidade Federal de São João Del-Rei. Tem experiência em Geocartografia e Ensino e Cartografia Digital.”

Ana Luísa Teixeira - “Graduada em Geografia pela Universidade Federal de São João del-Rei. Bolsista CAPES e Pós-graduanda no Programa de Pós-Graduação em Geografia da Universidade Federal de São João del-Rei. Foi líder do capítulo Unificar Ações e Informações Geoespaciais - UAIGeo UFSJ, vinculado à rede mundial YouthMappers, onde realiza projetos de mapeamento de comunidades ribeirinhas no município de Tefé - AM e na cidade histórica de São João del-Rei. 2021 Young Explorer National Geographic.”

Rafael Quetz Marques Oliveira - “Graduando em Geografia pela Universidade Federal de São João Del Rei, integrante do capítulo Unificar Ações e Informações Geoespaciais - UAIGeo UFSJ, da rede mundial Youthmappers, atuando com Mapeamento no munícipio de Tefé - AM e na cidade histórica de São João del-Rei.” 2023 2024 YouthMappers Leadership Fellow.’’

Ana C. F. Ribeiro

Laís Sessa (Iphan)

Hugo Frizzoni Candian - “Atualmente faço parte de uma iniciação científica da Universidade Federal de São João Del-Rei. Graduação em andamento na área de Geografia, com ênfase em bacharelado. Faço parte como membro da YouthMappers. e também como integrante do capítulo Unificar Ações e Informações Geoespaciais - UAIGeo UFSJ ”.

Ana Lara Santana de Almeida

Nathan Damas - "Possui Graduação em Engenharia Cartográfica e de Agrimensura, mestrado em Ciências Geodésicas e atualmente é doutorando em Ciências Geodésicas com foco em Cartografia, pela Universidade Federal do Paraná (UFPR). Membro fundado do primeiro capítulo da rede Youthmappers no Brasil, o Mapeadores Livres UFPR, e atual Embaixador Regional Youthmappers no Brasil. Participou do 2023 2024 YouthMappers Leadership Fellow".

Connect with the authors: @uaigeoufsj @youthmapperslatam

References

Oliveira, R. Q. M., Ventorini, S. E. (2023). MODELAGEM DE ÁREAS SUSCETÍVEIS ÀS INUNDAÇÕES NO SETOR COLÔNIA DO MARÇAL SÃO JOÃO DEL-REI, MINAS GERAIS. Revista Territorium Terram, 6(Edição Especial 1), 159–168. Recuperado de http://www.seer.ufsj.edu.br/territorium_terram/article/view/5306. Acesso em: 02 ago. 2024.

ONU Brasil. Organização das Nações Unidas do Brasil. A Agenda 2030. 2015. Disponível em https://brasil.un.org/pt-br/sdgs. Acesso em 06 de ago. 2024.

ZACHARIAS, Andrea; TEIXEIRA, Ana Luisa; VENTORINI, Silvia Elena; FERREIRA, André Barbosa Ribeiro; SANTOS, Thiago Gonçalves. A cartografia de síntese e as estruturas verticais e horizontais da paisagem em ambientes urbanos suscetíveis à inundação. Revista do Departamento de Geografia, São Paulo, Brasil, v. 41, n. 1, p. e177185 , 2021. DOI: 10.11606/eISSN.2236-2878.rdg.2021.177185. Disponível em: https://www.revistas.usp.br/rdg/article/view/177185 . Acesso em: 2 ago. 2024.