First two geothermal wells launched in Ladakh, India

Weather forecasts are essential for guiding the daily operations of various global sectors, such as airlines, power suppliers, farmers, and emergency services. However, experts warn that the increasing incorporation of artificial intelligence (AI) into meteorology, coupled with the growth of betting markets based on climate predictions, is raising the danger of manipulating the data used to generate these prognoses.

Impact on vital sectors

Climate projections go far beyond indicating the need for an umbrella; they form the basis for decisions with great economic and social relevance. For example, farmers use this data to determine the type of crop, the ideal time for fertilization, plan irrigation investments, and manage the grazing time of livestock.

Electric sector companies rely on these forecasts to decide the location of solar and wind farms, and even to set the prices of energy traded in the wholesale market. Additionally, forecasts are crucial for issuing warnings about extreme weather events and triggering disaster responses. More recently, they have acquired financial value in a new niche: prediction markets, where individuals bet money on real occurrences, including weather conditions.

Incentives for fraud in betting markets

The possibility of financial profit in these markets generates a strong incentive to tamper with meteorological data. Forecasts depend on accurate measurements of current atmospheric conditions, obtained from stations located at airports, power facilities, transport systems, and other points. Conventional models, such as the Weather Research and Forecasting Model (WRF) and the Integrated Forecasting System (IFS) from the European Centre for Medium-Range Weather Forecasts (ECMWF), combine these observations with physical models to project the future of the atmosphere.

Although equipment failures or device swaps can cause temporary interruptions in records, such problems are usually identified in real time or corrected later. Traditional systems employ a method called data assimilation, which compares each new reading with what the physical models predict and with data from neighboring stations, helping to detect inconsistencies.

Alert raised at French airport

Despite existing protective mechanisms, new threats have emerged. At the beginning of this year, media outlets reported that the meteorological station at Charles de Gaulle Airport in Paris had its records altered to show suspicious temperature peaks on April 6th and 15th. Authorities suspect that someone may have used a portable hairdryer or a lighter to artificially raise the sensor.

This manipulation resulted in high payouts for bettors who predicted a temperature of 22°C, since the actual average recorded on those days was around 18°C. According to the analysis, one participant received a prize of US$ 20,000 (R$ 102.5 thousand). The case was discovered fortuitously by members of a French non-profit climate organization, who noticed the anomalies and alerted the authorities.

Difficulty in detecting coordinated fraud

The authors point out that modifying the data of a single station is relatively easy to identify through human monitoring or established statistical methods. The real challenge lies in a scenario without human surveillance or where an intruder discreetly alters the records of multiple stations simultaneously. In this context, small individual changes may seem credible in isolation, complicating the identification of an orchestrated action.

Another obstacle is the time factor. Meticulous verification of metadata and data can take hours or days, while forecasts must be released within strict deadlines.

AI intensifies data dependency

The analysis highlights that the progressive adoption of AI in meteorology makes this problem even more critical. So-called data-driven models intrinsically depend on the quality of meteorological observations. Researchers at ECMWF are investigating the feasibility of generating high-quality forecasts directly from raw data, eliminating the data assimilation phase, which currently acts as a quality filter.

Other research groups are going further, working on integrating geospatial data, including information from weather stations, with large language models and autonomous AI systems to assist in real-time decision-making during storms and other severe events. Although these technologies promise greater agility, efficiency, and accuracy, the exclusion of human intervention in the process expands the spectrum of risks.

Scope of the threat: from fraud to national security

Experts outline different levels of threat. The simplest scenario involves a bettor manipulating a weather station to profit from prediction markets, as happened at Charles de Gaulle Airport. On a second level, groups of financial operators could alter forecasts related to renewable energy production to influence electricity wholesale market prices and gain economic advantages.

The most serious scenario would involve a state or sabotage agent manipulating data from one or more stations to trigger false alerts or prevent early warning systems from functioning when they are truly needed. In this situation, the problem transcends mere financial fraud, constituting an issue of disaster preparedness and potentially national security.

Three recommendations to protect AI forecasts

To mitigate these risks, experts propose three central actions. The first is to strengthen the security of weather stations, which implies increasing continuous monitoring, implementing automated anomaly detection systems, and maintaining human supervision. They also suggest that procedures for correcting meteorological records should become more agile to allow for near real-time fault identification.

The second recommendation focuses on strengthening the protection of data used by AI systems. This includes implementing distributed defenses across the entire processing chain, as well as tools capable of explaining model operation and increasing their resilience against targeted attacks.

The third measure consists of promoting shared responsibility among all parties involved in the process. Since data flows through station operators, national meteorological services, and forecasting centers, the authors argue that no single participant can ensure the integrity of the information in isolation. Therefore, any detected anomaly must be promptly communicated to all members of the chain.

The case as a warning sign

In the experts' assessment, the incident recorded at Charles de Gaulle Airport was detected in time, but it should serve as a warning for the future. As observational data gains prominence in modern weather forecasting systems, it becomes imperative to reinforce oversight mechanisms, improve coordination between competent bodies, and develop more robust methods to protect both the data and the AI models applied to meteorology.

Researchers are projecting that the current El Niño phenomenon could become the most intense ever documented. If these forecasts prove correct, the event could lead to record ocean temperatures, intensification of extreme weather in various parts of the globe, and an increased risk of humanitarian crises, especially in more vulnerable nations.

Analysis of Climate Models

Zeke Hausfather, a researcher at Berkeley Earth and author of the IPCC's Seventh Assessment Report, points out that climate models assign about a 90% chance that the 2026-2027 El Niño will be the most powerful since reliable records began. The study suggests that temperatures in the tropical Pacific Ocean area could rise by approximately 3.6°C above average, far exceeding previous records. However, scientists warn that it is still premature to define all the effects of the phenomenon with exactness.

Accelerated Strengthening of the Phenomenon

El Niño, which was officially declared on June 11, represents the warm phase of a natural climate cycle that unfolds over several years. This phenomenon is known for raising the global average temperature and promoting extreme events such as droughts, floods, and heatwaves. According to projections from the World Meteorological Organization (WMO), the event is expected to rapidly evolve into a strong condition between July and September. Furthermore, the NOAA Climate Prediction Center estimates a probability greater than 80% that the phenomenon will reach the very strong category by the end of the year, positioning this El Niño among the largest ever observed.

Emily Black, Professor of Terrestrial Processes and Climate at the University of Reading and senior researcher at the UK's National Centre for Atmospheric Science, emphasizes the importance of these projections. She told Live Science that the forecasts are impressive. While acknowledging that El Niño predictions always carry uncertainties, she stresses that the high consensus among models and the already visible warming in the tropical Pacific require the scenario to be treated with great seriousness.

According to Black, an extremely strong El Niño considerably raises the chance of destructive weather events occurring. She highlights that a very strong El Niño would substantially alter the probabilities of harmful weather events in many parts of the world, with a special impact on the Global South and its livelihoods.

The International Rescue Committee has also expressed concern about the worsening of the phenomenon. The organization indicates that the predicted conditions could trigger severe droughts and floods in regions of East Africa and Asia, affecting highly vulnerable communities.

Interaction with Global Warming

During the El Niño period, warmer waters accumulate in the eastern part of the Equatorial Pacific Ocean, which modifies atmospheric circulation and shifts the jet stream southward. In the United States, this typically generates warmer and drier conditions in the Northeast, while increasing the risk of flooding on the Gulf Coast and in the Southeast. On a global scale, the additional warming of the oceans adds more heat to the atmosphere, compounding the temperature increase caused by anthropogenic climate change.

Black underlines that the current context makes the phenomenon even more alarming. She explains that although El Niño is a natural process, it is occurring on an already warmer planet. This is relevant because a strong El Niño releases heat and energy into an already warmed environment. However, she warns that the impacts cannot be attributed solely to El Niño or only to climate change; the two factors interact. El Niño can potentiate droughts in certain areas, floods in others, marine heatwaves, monsoon disruptions, and exceptionally high global temperatures. Climate change, in turn, makes heat extremes more severe and can intensify heavy rainfall, as a warmer atmosphere retains more moisture.

Surpassing Previous Records

Hausfather reinforced the projections after an update to climate models in July. This analysis integrated 667 simulations from 14 different seasonal models. The results suggest not only a high probability that the phenomenon will be the most intense ever recorded but also that it may surpass the previous record by a notable margin. He stated that, with the July simulations, it seems that this year's El Niño has a high chance of being the strongest event since the beginning of reliable records, potentially being the strongest by a truly impressive margin.

NOAA defines El Niño conditions when the temperature in the tropical eastern Pacific region remains at least 0.5°C above the historical average, accompanied by changes in winds, atmospheric pressure, and precipitation. The phenomenon is classified as weak, moderate, strong, or very strong. Events with temperatures exceeding 2°C above average informally receive the name 'super El Niño,' although this is not an official scientific classification. The July models indicate that the anomaly could exceed 3.5°C by the end of the year.

According to Hausfather, the median of the projections indicates a peak of about 3.6°C, which represents approximately 0.8°C above the record set during the 2015-2016 El Niño, when the anomaly reached 2.75°C. About 91% of the simulations indicate that the current event will surpass the 2015-2016 record.

Scientific Caution is Necessary

Despite the robust projections, Black insists that the models are mere estimates and not guarantees. She stated that it is plausible for this to become a record-breaking El Niño, and recent forecasts make this possibility real, not distant. However, she advises caution when treating any probability estimate as a settled fact. She justifies this prudence by citing two points: first, the event has not yet reached its peak, and El Niños generally reach their maximum intensity closer to the end of the year; second, the definition of 'the strongest ever recorded' depends on the index, the dataset, and the reference used.

Historical Impacts of El Niño

Intense El Niño episodes have generated significant consequences in various parts of the planet. The 2015-2016 event, for example, was linked to a record hurricane season in the Central North Pacific, as well as severe droughts in the Caribbean and Ethiopia, and exceptionally high global temperatures. If current projections materialize, the phenomenon could equal or surpass both this episode and the notorious super El Niño of 1877-1878, which occurred before modern records. This historical event likely contributed to an extreme drought that fueled the Great Famine of 1876 to 1878, resulting in the death of over 50 million people. Researchers, however, point out that this tragedy was exacerbated by colonial agricultural policies of the time, demonstrating that humanitarian crises arise from the combination of environmental and socioeconomic factors.

Risk to Global Food Security

Even with the profound transformations that have occurred since the 19th century, experts warn that a super El Niño could still cause strong repercussions on global food systems. The concern is greater given that researchers point out that food insecurity is no longer limited to low-income countries, and climate change is putting pressure on agriculture in multiple regions. The climate cycle known as El Niño-Southern Oscillation (ENSO) alternates between the El Niño and La Niña phases every two to seven years, with a typical duration of nine to twelve months. According to Carbon Brief projections, 2026 should be the second hottest year ever recorded, and the strengthening of El Niño raises the probability of 2027 becoming the hottest year in history.

Black concludes that the record intensity of the phenomenon is secondary to its effects on the population. She states that even if the phenomenon does not set a new absolute record, its consequences can still be severe. Finally, she observes that these forecasts are worrying, but also useful, as they give societies time to anticipate potential impacts and act before the worst effects manifest.

The project presented by the Santa Marta Student Center in collaboration with the Mazzanti Team aims to create an architectural space that deeply integrates with the site, appropriating the environment and dissolving among the trees, recognizing the existing vegetation.

Architectural and Landscape Concept

The space's roof was designed to merge with the landscaping, resulting in an environment reminiscent of a grove. In this concept, the architecture not only accompanies but also enhances the natural characteristics of the terrain.

This space was conceived as a place for meeting and interaction, where activities such as resting, dialogue, and playing are presented as complementary methods of learning. Students have the opportunity to explore the outdoor environment in various ways, whether relaxing, swinging, climbing, lying down, or jumping, all immersed in nature.

Functionality and Structure

To support the dynamics of the location, sound and interaction devices were incorporated, which serve both to host events and to create ambient sound and act as a communication channel between students, reinforcing the participatory nature of the project.

The proposal includes a flexible structure, designed to adapt to the existing trees and operate as a network capable of growing and adjusting to the specific conditions of the site. Shade, considered an essential meeting point in the city, gave rise to a space that is simultaneously airy, shaded, and pleasant.

Centrality for the Community

The Student Center organizes various zones designated for studying, meetings, relaxation, meals, and leisure. Furthermore, it incorporates a module containing restrooms and a cafeteria, establishing itself as the main convergence point for the entire student community.