TL;DR:
- The Triunfo do Xingu region of the Brazilian Amazon suffered the equivalent of 700 football fields of deforestation in just one month in March 2023.
- PrevisIA, an AI platform developed by nonprofit environmental Imazon, has identified this area as the APA at the highest risk for further deforestation.
- Imazon is working to prevent deforestation through partnerships with local authorities.
- PrevisIA uses a two-pronged approach, taking into account trends in the region and factors that inhibit or contribute to deforestation.
- Monitoring the construction of roads, which are significant contributors to deforestation, is crucial to predicting and preventing it.
- The success of PrevisIA depends on who uses it and how effectively it is utilized.
- Imazon has established partnerships with state prosecutor’s offices in the region to prevent environmental crimes.
- The accuracy of PrevisIA has been exceptional, and the team at Imazon is continuously working to improve the model.
- The goal is for the tool to be so effective in preventing deforestation that it will no longer be necessary.
Main AI News:
The March of 2023 saw a devastating loss in the Triunfo do Xingu region of the Brazilian Amazon, with the equivalent of 700 football fields being cleared in just one month. This Environmental Protection Area (APA) spans over 16,000 square kilometers and is considered one of the largest conservation areas in the world. However, a new AI platform, PrevisIA, developed by nonprofit environmental Imazon, has identified Triunfo do Xingu as the APA at the highest risk of further deforestation.
According to PrevisIA, 271.52 square kilometers of forest in the conservation area are expected to be destroyed by the end of 2023. Despite being protected under Brazilian law, illegal activities such as mining, logging, and land-grabbing have already had a significant impact on the area, home to endangered species like the white-cheeked spider monkey and the jaguar.
Imazon is now working to establish partnerships with local authorities to prevent deforestation from happening in the first place. The destruction of the Brazilian Amazon is reaching alarming levels, with deforestation in March 2023 tripling compared to the same month last year. The first quarter of 2023 saw 867 square kilometers of rainforest destroyed, the second-largest area to be cleared in the past 16 years.
The idea for PrevisIA arose in 2016 when the Imazon team analyzed data from their Deforestation Alert System (SAD). Frustrated by receiving notifications after the damage had already been done, they asked if it was possible to create short-term deforestation prediction models.
The existing models at the time were long-term, looking at what would happen in decades, but PrevisIA was designed to get ahead of the destruction. Carlos Souza Jr, Senior Researcher at Imazon and Project Coordinator of PrevisIA and SAD, explains that the tool was needed to address the immediate threat to the Brazilian Amazon.
Souza and his expert team, consisting of a computer engineer, a geostatistics consultant, and two researchers, set out to develop an annual prediction model that takes a two-pronged approach. The model leverages trends in the region, using geostatistics and historical data from the government’s Prodes system for monitoring deforestation in the Amazon.
It also considers factors that inhibit deforestation, such as protected Indigenous and quilombola communities and terrain that is unsuitable for agricultural expansion, as well as factors that contribute to deforestation, such as high population density, settlements, rural properties, and dense road infrastructure.
According to Imazon, 90% of deforestation is concentrated within 5.5km of a road, making monitoring the construction of these roads crucial to predicting and preventing deforestation. The researchers used to manually analyze thousands of satellite images to detect new roads, but with PrevisIA, an AI algorithm automates the mapping process, allowing for quicker analysis and more frequent updates.
The team at Imazon partnered with Microsoft and Fundo Vale in 2021 to acquire the cloud computing power required to run the AI algorithm for mapping roads. This technological advancement has always been a key factor in controlling deforestation, and PrevisIA represents the next step in incorporating technology into the fight to protect the Amazon. Juliano Assunção, Executive Director of the Climate Policy Initiative and Professor at the Pontifical Catholic University of Rio de Janeiro, recognizes the significant potential of PrevisIA as a natural evolution in this process.
Souza and his expert team, consisting of a computer engineer, a geostatistics consultant, and two researchers, set out to develop an annual prediction model that takes a two-pronged approach. The model leverages trends in the region, using geostatistics and historical data from the government’s Prodes system for monitoring deforestation in the Amazon.
It also considers factors that inhibit deforestation, such as protected Indigenous and quilombola communities and terrain that is unsuitable for agricultural expansion, as well as factors that contribute to deforestation, such as high population density, settlements, rural properties, and dense road infrastructure.
According to Imazon, 90% of deforestation is concentrated within 5.5km of a road, making monitoring the construction of these roads crucial to predicting and preventing deforestation. The researchers used to manually analyze thousands of satellite images to detect new roads, but with PrevisIA, an AI algorithm automates the mapping process, allowing for quicker analysis and more frequent updates.
The team at Imazon partnered with Microsoft and Fundo Vale in 2021 to acquire the cloud computing power required to run the AI algorithm for mapping roads. This technological advancement has always been a key factor in controlling deforestation, and PrevisIA represents the next step in incorporating technology into the fight to protect the Amazon. Juliano Assunção, Executive Director of the Climate Policy Initiative and Professor at the Pontifical Catholic University of Rio de Janeiro, recognizes the significant potential of PrevisIA as a natural evolution in this process.
The success of PrevisIA will largely depend on who uses it and how effectively it is utilized. The AI platform, created by environmental nonprofit Imazon, has the potential to benefit a wide range of entities, including government agencies responsible for protecting the rainforest, as well as banks, investors, and product buyers. Imazon has established official partnerships with a few state prosecutor’s offices in the region, with the aim of using PrevisIA to prevent environmental crimes and irregularities rather than solely punishing them.
José Godofredo Pires dos Santos, a public prosecutor in Pará and coordinator of the environmental operational support center, recognizes the importance of proactive measures to protect the Amazon rather than reactive ones. The state prosecutor’s office in Acre also hopes to use PrevisIA to monitor high-risk areas and hold property owners accountable for any deforestation that occurs.
The accuracy of PrevisIA has been exceptional, with 85% of deforestation alerts within 4 km of the predicted location and over 49% of alerts in high or very high-risk areas, according to Souza, the senior researcher at Imazon and project coordinator of PrevisIA. The team at Imazon is continuously working to improve the model, and Souza hopes that one day, the tool will be so effective in preventing deforestation that it will no longer be necessary.
Conlcusion:
The development of PrevisIA by environmental nonprofit Imazon presents a unique opportunity for a wide range of entities to proactively address the ongoing deforestation of the Brazilian Amazon. The AI platform’s ability to predict and monitor high-risk areas with exceptional accuracy will be crucial in preventing further destruction of the rainforest and its valuable ecosystems.
This advancement in technology could have a significant impact on the market, as it offers a reliable and efficient solution for monitoring deforestation. Banks, investors, and product buyers will be able to make better-informed decisions based on the data provided by PrevisIA, leading to a reduction in the negative impact on the environment and potentially driving sustainable business practices. The partnerships established by Imazon with local authorities, including state prosecutor’s offices, demonstrate the potential for PrevisIA to be widely adopted and used effectively in the fight to protect the Amazon.
