A study using satellite imagery and machine learning techniques shows that many deforestation hotspots are outside the 11 municipalities currently controlled by Brazil’s federal government. Author: Agência Brasil

Using a method based on satellite imagery and artificial intelligence, Brazilian researchers have shown that the priority area for action to combat illegal deforestation may include an area 27.8% smaller than the 11 municipalities controlled by the federal government under the current a strategy known as the Amazon Plan. 2021/2022. This monitoring ignores the new deforestation limits outside the target areas.


According to an article by researchers published in June in Protective sheetsareas of the Amazon classified as high priority for having the highest deforestation The area was 414,603 km²2) this year, while the total area of ​​the plan for 11 municipalities is 574,724 km2. In other words, the observation area can be shortened by 160,000 km2which is about the size of Surin.

However, while the deforestation hotspots identified by the researchers account for 66% of the average annual deforestation rate, the 11 municipalities targeted by the plan account for 37% of the deforestation rate over the past three years (2019-21).

In the paper, scientists associated with Brazil’s National Institute of Space Research (INPE) and universities in the United States conclude that the proposed method will provide a tighter focus on monitoring and law enforcement. Furthermore, they point out, it opens up new frontiers of deforestation outside the priority area and thus not covered by the official monitoring plan.

“Using this new approach, we concluded that prioritizing areas with higher levels of deforestation would be more effective than limiting monitoring to specific municipalities. This is an important finding, given that the agencies responsible for law enforcement in this case, mainly IBAMA and ICMBio, have had their budgets and staff steadily cut. Some of these deforestation hotspots are in 11 municipalities, but others are nearby and are new frontiers,” Guillermo Augusto Verola Matavelli, corresponding author of the paper, told FAPESP. Matavelli is a researcher at INPE’s Earth Observation and Geoinformatics Unit. .

The National Council for Legal Amazonia (CNAL), which oversees the Amazon Plan for 2021/2022, responded to Agência FAPESP’s request for comment as follows: “The objective [of the plan] was to focus on where the emergence of illegal environmental activities most affected Brazil’s results environmental management without neglecting the need to act in other areas of the Legal Amazon.”

Legal Amazonia is a territory of more than 5 million km2 includes the states of Accra, Amapa, Amazonas, Maranhão, Mato Grosso, Paro, Randonia, Raraima, and Tocantins. It was created under the federal laws of 1953 to promote the special protection and development policy of the region.

According to CNAL, “11 municipalities were chosen because they had the largest deforested area and the highest number of fires, with the possibility of including others to be mapped by the Management and Operations Center of the Amazon Protection System. [Censipam].”

The board also said that INPE was one of the “leading institutions in the priority selection process” and that the scientists conducting the research “could provide institutional input when the opportunity arose.”

“CNAL always works with official information that is managed, processed and analyzed by official government agencies,” the statement said.

Advances in Data Processing

The authors of the article note that deforestation in the 11 municipalities covered by the plan has been significant in recent years, and that this is a reason for monitoring, but not enough to prioritize only these areas, which are as follows: São Félix do Xingu, Altamira, Nova Progresso, Pacaja, Portel, Itaituba and Ruropolis (Pará); Apui and Labreia (Amazons); Colniza (Matu Grosso); and Porto Velho (Rondonia).

They also note that despite the focus on these areas for monitoring and enforcement purposes, deforestation increased by 105% between February and April 2021 compared to the average for the same period from 2017 to 2021. .DETER, Brazil’s official deforestation warning program, pointed to 524.89 km2 new places of deforestation in these areas.

“The study confirms the importance of INPE, which for 60 years has trained outstanding researchers, producing science and technology from satellite data for the development of society and the country. Advances of Art data processing “Implementing the use of artificial intelligence to plan actions to combat deforestation is critical to mitigating the country’s environmental problems and building a national plan for sustainable development,” said Luis Aragao, the last author of the paper. Aragan heads the Department of Earth Observation and Geoinformatics of INPE. ,

Priority directions

Data sources for the study included INPE’s Amazon Legal Deforestation Satellite Monitoring Service (PRODES), which produces annual deforestation statistics used by the Brazilian government to develop public policy in the region. PRODES focuses on cutting and burning rates and has used the same methodology since 1988.

According to the latest report, the area of ​​cleared forests in the region was 13,235 km2 from August 2020 to July 2021. Compared to the same period last year, this increase was 22%, the largest since 2006.

“The idea for the article came about in February 2021, when Amazon’s 2021/2022 plan was announced,” Matavelli said. “Deforestation in 11 municipalities is said to account for 70% of the total deforestation found in the Amazon, but the PRODES figure was different. As we improved the model, we found it to be a useful tool to focus on monitoring and Enforcement more effectively.”

To identify priority areas, the researchers first defined what they call grid cells measuring 25 km by 25 km and regularly distributed across the Amazon. Using the Random Forest machine learning algorithm to predict next year’s deforestation hotspots based on sets of multivariate regressions, they placed each cell into a high, medium, or low priority class. According to the article, the method identified a greater proportion of areas at risk of deforestation, in terms of total area and public areas, where logging is illegal.

The model considered five predictors: deforestation in previous years, distance to grid cells with high cumulative deforestation in previous years, distance to infrastructure such as roads and waterways, total protected area in grid cells, and number of active fires.

The three priority classes were based on predicted deforestation, with values ​​below the 70th percentile classified as low, values ​​between the 70th and 90th percentile as medium, and values ​​above the 90th percentile as high. Grid cells classified as high were used for the 2022 priority area map with a total area of ​​414,603 km.2.

The authors also note that their method prioritizes action in border areas of 11 priority municipalities where deforestation is concentrated, captures other areas of increased deforestation not controlled by the plan, prioritizes based on land cleared in the previous year, and does not depend from geopolitical boundaries such as municipalities.

“Prioritizing these 11 municipalities will not be enough for Brazil to meet its international commitments, including the promise to reduce illegal deforestation to zero by 2028, announced at COP-26 [the 2021 UN Climate Change Conference]” – said Matavelli. “Moreover, the plan aims to reduce deforestation by 8,719 km2 per year, but the 2018 decree set a much lower target of 3,925 km2 in the year after 2020″.

This was a reference to Decree 9578 (2018), which consolidated the national policy on climate change and set the goal of reducing deforestation in the Amazon by 80% compared to the 1996-2005 average. This is one of the actions that Brazil is trying to curb greenhouse gas emissions.

In addition to its pledge to achieve zero deforestation by 2028, Brazil also announced at COP-26 that it will halve its greenhouse gas emissions by 2030 compared to 2005 levels and achieve climate neutrality by 2050. Increasing deforestation in the Amazon contrasts with these promises, with about 11% of greenhouse gas emissions coming from improper forest and land use, including deforestation and fires.

When Amazon’s 2021/2022 plan was announced, experts criticized the targets as insufficient because they were based on the average level of deforestation for the period 2016-2020, which was already 35% higher than the average of the previous ten years.

A call to further action

The article argues the need for a number of additional actions to combat deforestation in addition to direct methods of setting state policy goals. They should include environmental education and raising awareness, identifying and prosecuting actors who violate environmental laws and profit from illegal deforestation, promoting projects that invest in the green economy and forest conservation, and regulating public and indigenous lands.

“We used open source code to build the model and identify priority areas,” Matavelli said. “We are talking to the Terra Brasilis platform with the aim of including these areas in information available to all those who want to access it, so that it can be used in practice by any interested state or municipal authorities.”


Record deforestation in the Brazilian Amazon in February


Additional information:
Guillermo Matavelli and others. Scientific planning can support law enforcement efforts to stop deforestation in the Brazilian Amazon, Protective sheets (2022). DOI: 10.1111/conl.12908

Citation: Researchers Use Artificial Intelligence to Prioritize Action to Fight Amazon Deforestation (2022, August 24) Retrieved August 24, 2022, from https://phys.org/news/2022-08-ai-priority- areas-action-combat .html

This document is subject to copyright. Except in good faith for the purpose of private study or research, no part may be reproduced without written permission. The content is provided for informational purposes only.

Previous articleThe iconic image reappears on the 30th anniversary of Hurricane Andrew.
Next articleEmpire of Light: Cast, Trailer and Release Date