Researchers at UNIASSELVI developed a study that uses artificial intelligence and automation to compile climate data and assess the potential impacts of a Super El Niño event in Brazilian territory.
The investigation indicates a 97% probability of the phenomenon remaining active until 2027 and an 81% chance of reaching a very strong intensity, whose effects may vary depending on the location within the country.
How AI consolidates information for climate risk analysis
This study uses artificial intelligence to cross-reference data from various meteorological institutions, presenting projections on the dynamics of El Niño in a single analysis. The work was conducted by Fábio Henrique Moisés and Matheus Fernando Cunha, Software Engineering students at UNIASSELVI, under the supervision of Professor Pedro Sidnei Zanchett.
Additionally, the research is linked to Zanchett's RADIAN doctoral thesis, which focuses on decision support systems for natural disaster management, using generative AI, cloud computing, and automation.
It is important to note that despite the use of sophisticated technology, the study does not generate an autonomous climate forecast through artificial intelligence; its function is to organize and compare projections issued by both national and international meteorological centers. The novelty lies in the integration of AI, automation, and scientific knowledge to support decision-making in risk scenarios, given that climate change demands new ways of interpreting data.
Pedro Sidnei Zanchett, a professor at UNIASSELVI and advisor for the project, commented on the topic.
Diagnosis based on data from major institutions
To build this analysis, the researchers collected information from entities such as NOAA, NASA, INPE, INMET, CEMADEN, and IRI. The purpose was to confront different climate models and identify similarities between the forecasts. According to Moisés, the technology facilitates the transformation of existing data into more structured analyses.
The researcher stated: 'Artificial intelligence allows consolidating thousands of pieces of information produced by different institutions and transforming them into analyses capable of supporting faster and more informed decisions. The challenge is not to produce more data, but to connect the knowledge that already exists.'
Among the central aspects examined by the study are: the possibility of a high-intensity El Niño until 2027; the combination of data from Brazilian and foreign climate agencies; the use of AI and automation in organizing information; and support for planning against possible extreme events. Although the term Super El Niño is not an official technical nomenclature, it is frequently used to describe episodes of great magnitude of the phenomenon, caused by the unusual warming of the equatorial Pacific waters.
Impacts may vary according to the Brazilian region
The predicted effects will not be uniform across the entire country. While the North and Northeast regions may experience periods of greater drought, the Center-West and parts of the Southeast may record heatwaves and decreased precipitation. On the other hand, the South Region is pointed out as having the greatest potential for intense rainfall and flooding.
In the agricultural sector, the research signals possible losses to crops such as rice, wheat, soy, corn, and coffee. Furthermore, the combination of heat and altered rainfall patterns can increase thermal stress on livestock.
The study also highlights risks related to the expansion of fires in the Amazon, Cerrado, and Pantanal, coupled with the deterioration of air quality and the increase in respiratory and cardiovascular problems associated with heat peaks. Cunha reinforced that the research does not replace the work done by meteorological centers, stating: 'The consensus among the main international climate models is an important sign that the country must prepare itself. Our research organizes this evidence and demonstrates how intelligent technologies can contribute to increasing the response capacity in the face of extreme events.'
Even with technological advances, artificial intelligence remains a support tool. The results are conditioned by the quality of the analyzed data and may undergo modifications as the climate evolves, demonstrating how technology helps specialists organize information and anticipate potential risks, without replacing conventional scientific analysis.