As climate change supercharges fire weather across Southern California, predicting where and when the next wildfire will strike has become a matter of survival. Earth.Org trains machine learning models on historical weather patterns in an attempt to turn complex atmospheric variables into a powerful early-warning system capable of forecasting blazes before the first spark is struck. 

—

When examining climate-driven natural disasters in the United States, wildfires are frequently at the forefront of the conversation. As wildfires have become more prevalent, they have resulted in higher damages and impacts to society; in 2025, for instance, the Los Angeles wildfires swept through 40,000 acres of land in the area and racked up over $200 billion in estimated damage and economic loss.

But Los Angeles is just one of many examples of areas that are seeing a spike in climate change-fueled wildfires. According to Climate Central, the number of ‘fire weather’ days in San Diego, California, is on the rise; in nearby southeastern desert basins, including Palm Springs, it has increased by as much as 60 days between 1973 and 2024.

There are multiple variables that drive wildfire prevalence in these areas, but some of the most common ones include weather factors like humidity, temperature, and wind speed. For this analysis, we look at past data in San Diego to see if these specific variables can be predictors for the occurrences of wildfires in the area. 

The Data: Wildfire Weather Variables vs. Wildfire Occurrences

Historical wildfire data from the state of California (Cal Fire) as well as weather data from the National Oceanic and Atmospheric Administration (NOAA) between 2019 and 2025 were pulled to track wildfire occurrences and weather readings for San Diego during that time span. 

Daily weather readings for average humidity, average wind speed, maximum and minimum dry bulb temperatures, and peak wind speed were all combined to be used as multiple independent variables to predict whether a wildfire occurs for a specific day or not (yes or no) as the dependent binary outcome variable against all those readings. A map of wildfire occurrences recorded in San Diego between 2019 and 2025 is displayed below.

To predict wildfire occurrences for this data, three binary machine learning models were constructed and compared against each other to identify the model most suitable for predicting them.

To evaluate our models fairly, we split our data into two groups: 70% was used as a “training set” to teach the models the historical relationships between weather patterns and fires, while the remaining 30% was held back as a “testing set.” This test set acts as an exam, grading how accurately the models predict fires on data they haven’t seen before. 

Logistic Regression Model Results

The first of three models built was a logistic regression model – a machine learning algorithm used for binary classification problems to predict the probability of a binary outcome; in this case, predicting the probability of whether wildfires will occur or not. 

After the model was built, key parameters were found to decipher the results. The first was the accuracy score, which is used to assess the ratio of classification outcomes that were correct. The next parameter was the precision score, which is similar to the accuracy score, but focuses on the correctness of “clear” day predictions (number of days that did not have wildfires that were correctly predicted versus days that were predicted to not have wildfires but they actually happened). The results of the logistic regression model are displayed below.

The last parameter found was the confusion matrix, which displays where the model was right and wrong with predicted versus actual outcomes.

To visualize where our models succeeded and struggled, a confusion matrix was used. Rather than getting bogged down in data science jargon, think of this matrix as a simple scoreboard. It maps out four distinct outcomes: how often the model correctly predicted “clear” days (top left value), how often it successfully flagged an active wildfire (bottom right value), and the critical errors – the false alarms (predicting a fire that didn’t happen; top right value) and the missed warnings (failing to predict an active fire; bottom left value). 

Naive Bayes Model Results

The second of the three models built was a Naive Bayes classifier, which is another machine learning model that is used for binary classification problems (predicting two different outcomes). This model differs from logistic regression since it is used not only for calculating the probability of an event in the future, but it also assumes that the features (predictor variables) that are fed into the model are all independent of each other; in other words, the presence of one feature does not affect the other ones and they have no relationship to each other (hence why it is called a naive classifier). After this model was built, the same key parameters as the logistic regression model were found, with the results displayed below.

Random Forest Model Results

The last of three models built was a random forest, which is also utilized for binary classification problems but differs from the first two as it consists of a large number of decision trees, or links, that represent decisions. Each “leaf” represents an outcome (wildfire occurrence or not) that all work together collectively as an ensemble as each tree gives a classification prediction and whichever outcome tallies the most results in the ensemble will provide the prediction for the model. After this model was built, the accuracy and prediction scores were found again, with the results displayed below.

Final Thoughts

Going back to the original question of whether or not we can use potential wildfire forecasting variables to predict wildfire occurrences in San Diego, the results show that it can be done, with similar results for all three models demonstrating very high accuracy and precision scores. However, the random forest model displayed slightly higher values for accuracy (96.12% versus 95.96% and 95.80%) and precision (96.42% versus 95.96% and 95.95%) compared to the other two models. 

What’s more, looking at its confusion matrix, the random forest model had a slightly stronger set of predictions since, even though the random forest model predicted two false alarms for wildfire occurrences and the Naive Bayes classifier had one, the random forest model was able to detect the most actual wildfire occurrences in the data (three of them) while the other two models predicted zero. Even though we could see that most days do not have a wildfire occurrence in the data, there were still occurrences in about 4% of the total days of data. So, as the random forest model was able to correctly predict a few instances of those actually happening, that model would be preferable to utilize for this scenario. 

Even though the results for the analysis were strong, there are additional ways in which it can be expanded upon, such as investigating different regions in California with the same Cal Fire data or different states in the United States with different statewide data sources. Another way it could be expanded upon is by running additional machine learning algorithms suited for binary classification, such as a regular decision tree, a neural network, or a support vector machine (SVM), to use as additional comparison points to the three models constructed for this analysis.

As wildfires continue to increase in frequency in many regions due to climate change, looking into different methods to predict their occurrences provides important insights into proper associated preparations as well as driving sustainability initiatives.

Featured image: slworking2 via Flickr.

Is there a connection between cattle methane emissions and the total usage of different synthetic agricultural fertilizers in the United States? This is what the data says. 

—

When looking at the greenhouse gases that are driving climate change, one that is often brought up is methane (CH4). Methane is the second most abundant human made greenhouse gas after carbon dioxide (CO2) and it is 28 times more potent at trapping heat in the atmosphere. 

Methane is produced from a variety of anthropogenic sources, such as landfills, oil and natural gas systems, and coal mining, with agriculture being the largest source, accounting for 40% of global methane emissions. Cattle, which is widely consumed in Western diets and primarily in the US, accounts for the largest share of these emissions – around 75%.

Multiple factors drive these livestock emissions, but one overlooked variable is the role of synthetic fertilizers used to grow the massive amounts of feed crops required to sustain these cattle. In this analysis, we look at the historical data of synthetic fertilizer use in the US to see if it can act as a reliable predictor for the country’s cattle methane footprint. 

The Data: Fertilizer vs. Methane

To find a pattern, we pulled historical data from the Food and Agriculture Organization (FAO) spanning 1961 to 2023. We tracked the US agricultural consumption of three common fertilizer types: nitrogen (N), phosphate (P205), and potash (K20).

We compared these numbers against methane emissions specifically from beef cattle, which carry a much higher carbon footprint than dairy operations. Our goal was to see if changes in fertilizer use directly mirror shifts in methane output.

Before running a deeper statistical model, we plotted the raw data on scatter plots to visualize the trends. As seen in Figure 1, beef cattle methane emissions steadily increase alongside rising phosphate and potash usage. Interestingly, nitrogen fertilizer shows the opposite trend – suggesting that phosphate and potash have a much stronger, direct correlation with the US cattle footprint than nitrogen.

After the scatter plots were produced, the data was split into train and test sets, with 70% of the data being used to train it and 30% of the data used to test it. The train set is used to help the model learn how the data is set up and build a model with it. The test set is used to help assess the performance of the model with a less familiar data set than what was used to build it. 

The linear regression models were then built, and key parameters were found to help interpret the results. One of those is the r² value (coefficient of determination), which measures how well the model to predict methane emissions fits the data that is provided. Another parameter that was found is the mean squared error (MSE), which provides a way to assess accuracy by measuring how close the predictive linear model is to the actual data. The r² and MSE values for both the train and test data for the three fertilizer regression models are displayed below. 

As we can see in Table 1, the phosphate model had the highest overall r² value with the train data and the nitrogen model had the highest r² value specifically with the test data. There were some differing positive and negative correlations with the phosphate and potash models with the train and test data, but the train data (which represents a higher percentage of the data) resulted in a positive correlation between the two variables in all three models. For the MSE values, the nitrogen and potash models had higher values than the phosphate model which means the points in the data for the former model prediction lines are not as close to the latter (the phosphate model), but overall there is still some degree of correlation when we take the r² values into consideration. 

Regression lines were also plotted for the three models, with the train data highlighted in blue and the test data highlighted in green for each graph in Figure 2. We can see that the trend lines point upward for the phosphate and potash models, but the nitrogen line points downward, indicating that the higher variability in the methane emissions versus nitrogen data resulted in a slight downward trend compared to the other two fertilizers.

Final Thoughts

Addressing the original question of whether there is a relationship in the data between cattle methane emissions and the total annual usage of different synthetic fertilizers, the results demonstrate somewhat of a correlation, particularly for phosphate and potash. However, nitrogen fertilizer showed an opposite correlation and did not follow the expected upward trend, which could be due to a few potential causes. 

One possible cause is that methane emissions from cattle are typically more tied with their digestive process– known as enteric fermentation – in which microbes in the digestive system decompose food and produce methane as an output. According to the US Department of Agriculture (USDA), increasing efficiencies in fertilizers like nitrogen could lead to a reduction in methane from those processes. However, there is still no evidence to suggest that the total usage of those fertilizers directly correlates to total cattle methane emissions.

Another point to consider for the mixed results is that the greenhouse gas outputs of different synthetic fertilizers typically release different direct outputs, such as nitrous oxide (N2O), which is directly released from excess nitrogen fertilizer usage, rather than methane. Because potassium and potash fertilizers do not contain carbon or nitrogen in them, they typically do not release a large quantity of greenhouse gases as a direct byproduct, although they may release them indirectly through other lifecycle processes, such as in the processing and transportation of those fertilizers.

There are several ways in which this analysis could potentially be expanded upon, such as investigating additional high-emitting livestock like lamb and goat instead of cattle and different greenhouse gas outputs such as carbon dioxide and nitrous oxide instead of methane. Additionally, the usage of different fertilizer products such as ammonia and sodium nitrate could be examined to see if there are additional correlations to be drawn. 

As emissions of different planet-warming greenhouse gases from agricultural processes continue to increase, looking into these types of correlations provides key information to guide climate change mitigation efforts.

When the high-speed rail line between Barcelona and Madrid opened in 2008, hundreds of thousands of people ditched the plane in favor of the train, helping significantly lower carbon dioxide emissions in the country. Progress elsewhere in Europe has been slow. But with the European Commission’s new plan to accelerate the development of high-speed rail in Europe, the continent’s travel emissions could plummet.

—

It is a well-known fact that short-haul and domestic flights are the most polluting forms of travel. Despite this, European governments have long subsidised and invested in air travel over rail travel, resulting in today’s norms of continental travel being very plane-centric. 

In 2016, the European Union announced the Fourth Railway Package, a set of legislation aimed at “revitalising the rail sector and making it more competitive vis-à-vis other modes of transport.” While this package did not directly fund any infrastructure or route developments, nor subsidies for the rail industry, it removed a lot of the red tape that had historically held back rail development, creating a legislative environment more conducive to a competitive rail industry and paving the way for more concrete initiatives and investment.

Last month, the European Commission announced a comprehensive plan to accelerate the development of high-speed rail in Europe through greater coordination and streamlining of ticketing, timetabling, funding, operations and legislation. The plan aims to deliver a continent-wide high-speed rail network by 2040. This will be achieved through a series of actions, including improving funding sources and private investment coordination as well as cross-border rail ticketing and booking systems; simplifying train driver certification, and removing redundant national rules.

The project is also expected to significantly reduce the EU’s transport-related emissions, supporting its goal of achieving carbon neutrality by 2050. Spain’s high-speed rail impacts on CO2 emissions are a good example to illustrate the potentially huge impact that an Europe-wide initiative could have.

Case Study: Spain

In 2008, a new high-speed rail line opened between Barcelona and Madrid, cutting the journey time between the two cities by train from around nine to two-and-a-half hours. Almost instantly, air passenger numbers declined. 

Data from Eurostat shows a reduction of roughly one million passengers per quarter –  or 40% – immediately after the high-speed line was opened in February 2008, rising to around 60% by 2024. The Comisión Nacional de los Mercados y la Competencia (CNMC) – Spain’s independent competition regulator – has been only collecting data on passenger numbers since 2016.

A 2005 study at Oxford University calculated that around 153 grams of CO2 is emitted per passenger kilometer on short-haul flights. This means that every passenger that flies the 481-kilometer stretch between Barcelona and Madrid airports emits roughly 74,000 grams of CO2.

As for high-speed rail, the UK Department for Energy Security and Net Zero (DESNZ) and the French public train company SNCF estimate four and three-and-a-half grams of CO2 per passenger kilometer, respectively. The high-speed line between Barcelona and Madrid is 681 kilometers long, meaning every passenger emits around 2,550 grams of CO2.

The Results

Using the passenger figures from Eurostat and CNMC, we can draw a graph that shows the number of passengers by mode of transport over time. As we can see, the total number of passengers between Barcelona and Madrid wasn’t hugely affected, as the new train route took on the passengers that were no longer choosing to fly. This shows that peoples’ ability to travel between these cities was not affected – only they now had an alternative, which quickly became the preferred mode of transport.

But what is really striking about these figures can be seen when we apply our estimates for the amount of CO2 emissions per passenger. These estimates allow us to compare the CO2 emissions over time of air travel and train travel between Barcelona and Madrid.

Here we can plainly see just how much greener high-speed rail is than air travel. Even though the number of train passengers was higher than plane passengers from around 2012 onwards, rail emissions pale in comparison to the emissions from air travel. Had the high-speed line never been built, CO2  emissions from this route would be three times as great in 2024.

As more people opt to take the train instead of a flight, as has been the case since the pandemic, CO2 savings grow. In the second quarter of 2024, the actual total emissions were 275 million kilograms lower than if all passengers had flown, which is roughly equivalent to the emissions of over 183,000 British households over the same time period.

Learnings for Europe

The European Commission’s newly announced plan targets a number of routes. Some of the biggest improvements include reducing the Copenhagen–Berlin route from seven to four hours; Berlin–Vienna from over eight hours to four and a half; and Sofia–Athens from nearly 14 hours to just six. Additionally, the plan would create new routes, between Paris, Madrid and Lisbon, and between the Baltic states and Warsaw.

No doubt, the improvement, or opening, of high-speed rail routes across Europe will present an alternative to flying as a means to travel between two cities, and as a result will take passengers off planes and put them onto trains instead, lowering the overall emissions. As the Barcelona–Madrid route shows, this can have a stunning impact – across 2024, around one million fewer tonnes of CO2 were emitted than if the high-speed line had never been built, which is more than the total annual CO2 emissions produced by the Faroe Islands, Gambia or French Polynesia.

It is difficult to estimate exactly how large the impact could be on any given route. The corridor examined here was very successful, and a 60% reduction in air passengers is not to be taken for granted. Nevertheless, it is clear that investment in railways as an alternative to air travel has the potential to have a hugely positive environmental impact, and the EC’s plan must prioritize maximizing passengers in order to make this impact as big as possible. The key to this will be  through affordable tickets, fast and reliable journeys and reduced red-tape, and it’s vital that the EC recognises this. It’s high time that countries begin to offer the same, if not greater, legislative benefits that the aviation industry has enjoyed for decades.

Featured image: Wikimedia Commons.

Kazakhstan has lost 21% of its per capita water availability since 1999. But what are the main forces behind the country’s rapidly worsening water crisis? 

—

Only 42% of Kazakhstan’s water is available for consumption due to outdated infrastructure and poor management. Over 45% of available water is transboundary, meaning it comes from sources such as the Irtysh and Ili rivers from China and the Syr Darya from neighboring Kyrgyzstan and Uzbekistan, making Kazakhstan vulnerable to politics and foreign water mismanagement.

The poor irrigation productivity, at less than US$0.5 US per cubic meter, means that for every cubic meter of water (1000 liters) used for irrigation, the resulting agricultural output generates less than half a dollar in value. Combined with unmaintained infrastructure and low returns from water-intensive crops this also plays a role, translating into huge missed economic opportunities. The inadequate sanitation alone costs the government some US$750 million per year, compared to $500 million in investments needed to close this gap.

Glacial Threat

Kazakhstan’s water sources are heavily dependent on glacier-fed rivers from the Tien Shan mountains. Glaciers in this region have already lost about 30% of their mass and may retreat by another half by 2050.

The initial deluge in meltwater is temporary; soon enough, these rivers will dry out in the summer months, when water is most needed for cooling and irrigation, eventually leading to seasonal shortages and ecological collapse.

Agriculture: Thirsty and Wasteful

In 2021, 63% of annual water withdrawal in Kazakhstan was spent on agriculture, making it the largest sector of water waste, compared to 19% for domestic use. It is reported that only half of the water determined for agricultural use is utilized effectively or reaches the crops.

Additionally, only 16% of all irrigated land uses modern irrigation techniques, such as drip or sprinkler systems. The government introduced a policy on Water Resources Management (2023–2029), which acknowledges these challenges and aims to reduce agricultural water loss through main canals from 20% to 15% by 2029.

The agricultural inefficiencies not only stress the country’s water resources but also put additional pressure on food security and economic stability.

Unequal Access to Water

Despite Kazakhstan’s efforts to improve water infrastructure, differences between urban and rural areas still exist. While urban centers enjoy access to centralized water and sanitation, rural residents face challenges that affect their health and quality of life.

As of 2024, around 99.5% of urban residents had access to clean water, with 97.8% of rural settlements having similar access. Access to piped water shows a more pronounced discrepancy. While 94% of urban households are connected to piped water systems, the number is closer to 60% for rural households – forcing people to rely on other sources of water like public standpipes or wells.

The differences in water access between urban and rural areas are not only an infrastructural issue; they also have serious implications for public health, economic development, and social equity. Addressing these challenges requires investments in rural water infrastructure, community engagement, and policy reforms targeted at ensuring equal access to water sources across Kazakhstan.

Geopolitics

Kazakhstan’s water resources, as was mentioned earlier, are dependent on transboundary rivers – particularly the Ili and Irtysh rivers, which originate in China and are vital for Kazakhstan’s agriculture, industry, and ecosystem in general.

Negotiations between Kazakhstan and China to reach an agreement on water equitability have been ongoing. But despite some progress, such as the joint construction of hydraulic facilities on shared rivers like the Khorgo, a broad agreement on water allocation remains ambiguous.

The situation is complicated by Kazakhstan’s dependence on other transboundary rivers such as the Syr Darya, which flows through Kyrgyzstan and Uzbekistan before reaching Kazakhstan. Effective use of these water resources requires regional cooperation, which is challenging due to differing national interests and priorities.

Kazakhstan has taken steps to reduce tensions with its neighbours. For example, Kazakhstan and Uzbekistan installed transboundary water meters to monitor water consumption and share data, improving cooperation and transparency.

However, the challenge remains because of the lack of binding international agreements and the ever-increasing demand for water in the region, depicting risks to Kazakhstan’s water security. Addressing the geopolitical challenges is essential for sustainable water management and preventing conflicts over this vital resource.

What the Future Holds

The predictive model Earth.Org has built depicts an uneasy reality. Renewable internal freshwater resources have been in decline since the 1990s, and forecasts show this trend will continue, potentially dropping to 2,750 cubic meters by 2031.

The widening uncertainty interval emphasizes the importance of action – policy changes, investments in water-saving technologies, and improved regional cooperation.

Water scarcity has further effects on public health, economic resilience, and national security. The best time to act is now.

You might also like: The Aral Sea Catastrophe: Understanding One of the Worst Ecological Calamities of the Last Century

Energy demand is growing in India, but current renewable additions are not enough to meet this rising demand, let alone replace existing coal. 

—

In Gujarat’s white desert, the Khavda renewable power plant is set to become the world’s largest clean energy installation, generating a staggering 30 GW of solar and wind power by 2030. Yet, across the country, a 3 GW coal-fired plant in Chandrapur, Maharashtra, continues to spew smoke, one among hundreds that still power over 60% of India’s grid.

This juxtaposition captures the central tension in India’s energy transition: How can a developing nation reconcile its growing hunger for energy with its increasingly ambitious climate goals?

Development vs. Decarbonization

Unlike developed economies, India is still growing. Its people still consume far less energy than the global average. Electricity provision – a given in the developed world – is still a development imperative here. 

Over the past two decades, India has made remarkable progress in electrification. Thanks to targeted government programs, electricity access rose from just 60% in 2000 to near-universal coverage today, and electricity shortfalls have fallen below 1%. However, this electrification has been driven primarily by coal. 

Coal capacity surged in the previous decades, and continues to rise, albeit more slowly. Today, the electricity sector accounts for over half of India’s emissions. While its per capita emissions remain low, India is now the world’s third-largest emitter of greenhouse gases.

Even so, India has emerged as a climate leader. Its 2015 Nationally Determined Contribution (NDC) pledged that 40% of its electricity capacity would come from non-fossil sources by 2030. In 2022, it raised that target to 50% and set a net-zero goal for 2070. Since 2015, India’s renewable capacity has grown nearly 20% annually, crossing 200 GW in 2024.

Still, as the world’s “well below 2C” window closes fast, attention is turning to India. While global emissions are expected to peak soon, India’s are still climbing. The country is projected to drive over a quarter of global energy demand growth through 2040. As the world is trying to phase out coal, why isn’t India able to move away from it? 

Renewables Are Cheaper, But that Is Not Enough

Over the past decade, the costs of solar and wind have plummeted. On a levelized cost of energy (LCOE) basis – the average unit cost of electricity over a plant’s lifespan – renewables are now cheaper than coal. Even by variable cost comparisons, solar now beats coal. Then why are new coal plants still being built and planned? 

The answer lies in what cost metrics do not capture.

A Kilowatt Isn’t Always a Kilowatt

Solar and wind – classified as variable renewable energy (VRE) – generate power only when the sun shines or the wind blows. A solar plant generates only 15-25% of its maximum output over a year – known as the Capacity Utilization Factor (CUF) – while wind averages 25-35%. A coal plant, in contrast, can run 24/7, with CUFs as between 70-90%. This means that much more VRE capacity must be built to meet the same demand, requiring greater upfront investments, especially challenging in a country where capital is expensive.

But the challenge isn’t just how much power is generated. It’s also about when and where.

Integration Challenges: Time and Space

A key challenge is timing. India’s power demand peaks twice daily: once at noon and again after sunset. Solar power helps with the former, but not the latter. And with solar making up the bulk of new renewable additions in India, there is little clean energy available after sunset, just when people return home and air conditioner use spikes. 

This temporal mismatch between demand and supply could be solved through battery storage. Solar output during the day could be stored and used for nighttime demand. But utility-scale battery storage remains far from commercially viable. Another option is to use flexible fuels like natural gas: its generation can be ramped up or down quickly to complement renewable generation. Countries in the Global North rely on it to complement VRE. India, however, lacks the infrastructure for gas at scale. Instead, it relies on coal.

Coal is less nimble, and costly to turn on and off. As a result, coal plants are often kept running all day, even when renewable generation is high. This lowers their CUFs, and drives up their per-unit costs, since fixed costs are spread over fewer units of output. Paradoxically, it was the overcapacity of coal built in the previous decade that provided much needed flexibility to integrate VRE at scale. But it also drove coal CUFs to record lows.

Geography is another challenge. 

While coal thermal plants can be built close to demand centres, India’s solar and wind resources lie far away. The Khavda facility, for instance, lies far from Gujarat’s industrial and urban hubs. Connecting these new energy sources requires massive new transmission infrastructure.

Moreover, India’s renewable potential and current installed capacity are geographically concentrated in six southern and western states. Meanwhile, coal-rich states in central and eastern India account for much less. And although India has a nationally integrated grid, actual dispatch and balancing are often done at the state level. So excess solar in Rajasthan cannot easily serve daytime peaks in neighboring states. 

While transmission infrastructure is being expanded, it is struggling to keep pace with renewable deployment. This mismatch has led to curtailment – clean power that goes unused because the grid cannot absorb it – which has increased in multiple states in recent years.

These integration challenges – temporal, geographic and institutional – add hidden costs. They lower CUFs and increase curtailment, neither of which are captured in LCOE calculations. They also limit the share of VRE that can be integrated to the grid, regardless of how cheap it becomes. The remaining demand must be met by other sources, which in India’s case, is predominantly coal. 

When Might Coal Peak?

Rahul Tongia, Senior Fellow at think tank Centre for Social and Economic Progress (CSEP), offers a useful framework: the Ladder of Competitiveness. 

• Stage 1: VRE is costlier than new coal → Coal dominates new capacity.
  
• Stage 2: VRE becomes cheaper than new coal → Renewables meet some new demand, but coal persists. India is here today.
  
• Stage 3: VRE + storage becomes cheaper than new coal → Renewables meet nearly all new demand, coal additions stop.
  
• Stage 4: VRE + storage is cheaper than existing coal → Renewables start to meet demand earlier met by coal. Coal starts to decline.

A number of studies project India’s annual electricity demand growth at 6-6.4% until 2030. According to CSEP, meeting this demand without accounting for storage would require adding 46 GW of solar and wind each year, consistent with India’s 500 GW target. But factoring in storage pushes the number higher.

CEEW finds that once transmission constraints are included, the 500 GW target may fall short, but diversifying VRE capacity across states could help reduce the shortfall. Ember and TERI find battery costs keep falling by roughly 7%, India may remain in Stage 2 at least until 2032, when storage at scale becomes viable. Until then, VRE could meet over 75% of solar-hour demand, but only a third in non-solar hours, the rest met by coal.

Yet, actual deployment is lagging. Annual additions may be lower–20-30 GW for solar and 4-8 GW for wind. If that gap persists, coal dependency may grow further. 

The Way Forward 

Until storage scales, increasing the flexibility of India’s power system is key. Demand response measures such as Time-of-day (ToD) pricing could incentivize customers to shift demand to times when the supply is more plentiful, reducing the need for battery storage. 

Grid operators need better real-time monitoring and control systems to manage intermittency, such as a sudden cloud cover over solar fields. Coal plant flexibility increases, such as ramping down during solar peaks and ramping up during time hours, could enhance supply flexibility, but doing so could be costly, making operators reluctant. Finally, transmission reforms, such as improving inter-state coordination could drastically reduce curtailment, cut emissions and minimise unnecessary coal additions.

Final Thoughts 

We began with a striking image: the gleaming Khavda solar park and the aging Chandrapur coal plant. But perhaps framing India’s energy story as a battle is misleading. 

India’s energy demand is still growing. Current renewable additions aren’t enough to meet this rising demand, let alone replace existing coal. So for now, it is coexistence – with coal filling gaps that renewables can’t yet close.

The stakes are global. How India balances its growth and decarbonization over the next decade will play a major role in determining whether the world can limit warming below 2C. But for now, the world may have to accept the fact that coal isn’t going anywhere anytime soon.

Earth.Org’s Data Analyst and Visualization Expert Madina Tussupova uncovers the patterns, causes, and real consequences of global temperature rise.

—

Instead of exploring average temperature data, this article looks at temperature anomalies; in other words, deviations from average temperatures. Why is it more useful? Because the “normal” temperature range for each location is widely different – by looking at anomalies, we can effectively measure the temperature trend while avoiding climate bias.

References to global warming in this article include both land and ocean warming. The decision to not overlook ocean warming makes sense, as 90% of global warming occurs in the ocean.

Historical Temperature Trends

Historical records from the mid-20th century indicate a steady rise in average temperatures. While natural climate variability has caused certain fluctuations, the general trend has been a consistent warming.

Our analysis show the following observations:

• The global average surface temperature has risen by approximately 1.5C since 1924.
• The global warming rate of this century (1924-2024) is approximately four times higher than the rate observed in earlier years (1850-1923), emphasizing a significant acceleration in climate change during the last century [linear regression was used to calculate the slopes]
• The 21st century has seen some of the warmest years on record, with 2024, 2023, and 2016 topping the ranking. 

Regional Differences in Temperature Rise

The rise and impact of global warming are not the same across the world. Some regions face more accelerated warming due to specific climate and weather conditions. 

The highlighted observations are as follows:

• The Arctic has warmed at a rate nearly four times the global average, leading to significant ice melt and ecosystem disruptions.
• Continental regions such as North America, Europe, and Asia have experienced significant increases in extreme weather events, such as compound droughts and heatwaves.
• Oceans have absorbed much of the excess heat, contributing to rising sea levels and more frequent and intense hurricanes.

Global Warming and Greenhouse Gases

Greenhouse gas emissions, particularly methane, nitrous oxide, and carbon dioxide (CO2), are the main driver of global warming. Human activities are the dominant force behind rising CO2 concentrations, accounting for 55.4% of total emissions. Industrialization and urbanization have also contributed significantly to localized warming, known as the urban heat island effect.

Natural emissions play a role, too, ranging from 18.13 to 39.30 gigatonnes of CO2 equivalent (Gt CO2-eq) annually, with a likely value around 29.07 Gt CO2-eq. Key natural sources of emissions include forest fires, oceans, wetlands, permafrost, volcanoes, and earthquakes.

CO2 levels reached approximately 420 parts per million (ppm) in 2023 – the highest value since records began. That year was also the warmest year within the last century, until 2024 – currently the hottest on record.

Deforestation and land-use changes have further reduced the planet’s ability to absorb CO2, intensifying the temperature rise. Forests, like oceans, are invaluable carbon sinks, absorbing and storing large quantities of CO2 from the atmosphere. 

Soil holds carbon in forms like permafrost or peat, but temperature rise makes it hotter and drier, making it more difficult to store more carbon.

In the oceans, one of the main players in carbon absorption is phytoplankton, microscopic bacteria that consume CO2 and release oxygen. However, high concentrations of microplastics in the water disrupt the structure of algal communities, which negatively affects the entire marine ecosystem. Plants absorb CO2 thanks to a process called photosynthesis. With the help of light, trees generate glucose – which they later use to produce cellulose and other things they need to grow – and oxygen, which they release into the atmosphere.

Predictions and Future Outlook

A polynomial regression model was used to make predictions for the next 25 years. While it fits the overall trend well, we can see slight underfitting around the years 2022 and 2023. This can be explained by a spike in anomaly that could not be captured by the model.

This means our predictions might be underestimated – in reality, the values could be much higher than forecasted. Nevertheless, we can still capture some valuable insights from the model’s predictions:

• If current patterns persist, we could see anomalies surpassing +2C by 2045, which crosses several critical climate thresholds.
• The model also predicts that we could breach the 1.5C anomaly as early as the 2030s – a major climate milestone referenced in the Paris Agreement.

Conclusion

Historical temperature time series affirm that, in the last century, global warming has accelerated, with much of this intensification attributed to anthropogenic activities. While the impact varies from place to place, what makes it truly alarming is the global nature of temperature rise.

The time left to halt greenhouse gas emissions, transition to carbon-free energy, and push for sustainable environmental policies is very limited.

Featured image: Arian Zwegers/Flickr.

The average wildfire season in Western US is now 105 days longer, burns six times as many acres, and sees three times as many large fires, according to Climate Central. Earth.Org looks at the data to establish whether there is a link between increased wildfire activity and drought severity.

—

Drought is one of the most devastating natural disasters, affecting millions of people worldwide. Unlike hurricanes or earthquakes, droughts develop slowly and can persist for months, years, and even decades, leading to lasting effects on agriculture, water supply, and ecosystems. 

In 2024, 48 of 50 US states faced drought conditions, the highest proportion ever recorded in the country, while wildfires scorched over 8.9 million acres nationwide. As climate change accelerates, droughts are becoming more frequent and widespread, and their relationship with wildfires is no longer speculation.

Drought Trends

Drought occurs when a region experiences prolonged periods of lower-than-average precipitation, leading to water shortages and drier lands. Both natural and human factors can influence it, including varying climatic conditions, rising global temperatures, deforestation, and unsustainable water use. 

The US Drought Monitor provides historical drought data, classifying drought severity into categories ranging from D0 (Abnormally Dry) to D4 (Exceptional Drought).

A time series plot from 2000 to January 2025 illustrates trends across different drought categories (D0 to D4), revealing a periodic pattern. Severe droughts (D2-D4) peak every 3-5 years, likely influenced by climate oscillations such as El Niño and La Niña. Autocorrelation tests confirm this periodicity, showing strong correlations at 2-3 year lags, particularly for D3–D4 droughts. 

Additionally, drought severity has intensified, with “exceptional drought” coverage increasing by 12% since 2000. While mild droughts (D0–D1) follow annual wet/dry cycles, severe droughts (D2–D4) persist for multiple years. 

Given this cyclical behavior, an important question arises: do wildfires follow a similar pattern? 

Wildfire Trends

Using historical wildfire data from the National Interagency Fire Center, we analyzed the total acres burned over the years. The data reveals that wildfires also exhibit a periodic pattern. This raises a critical question: is there a direct relationship between drought and wildfires?

To investigate this, we utilized a Normalized Weighted Drought Index, which quantifies drought severity on a scale from 0 (least severe) to 10 (most severe) by weighting different drought categories. Similarly, wildfire data including total acres burned and the number of fires was normalized to enable a direct comparison with drought severity. 

Normalization is a method that scales different types of data to a common range, allowing us to compare them more fairly. For instance, drought severity may range from 0 to 10, while acres burned could be in the thousands or millions. Without normalization, one might appear more significant just because it uses larger numbers. By bringing everything to the same scale, we avoid misleading comparisons and can better evaluate how closely drought conditions and wildfire activity are connected. This approach provides a dimensionless framework to explore their correlation more accurately.

The plot reveals a strong relationship between drought severity and wildfire activity, with extreme drought years, such as 2020, coinciding with an increase by 2.5 times in burned area compared to non-drought years. However, an anomaly appears between 2011 and 2015, where overall drought levels were lower, yet wildfire activity remained high. 

A more focused analysis considering only severe drought conditions (D3-D4) provides further clarity, highlighting that localized extreme droughts can significantly impact fire activity, even when total drought levels appear moderate. This suggests that while broader drought classifications may not always capture fire risk accurately, high-intensity droughts in fire-prone regions play a crucial role.

The data also shows that fire frequency increased by 40% in the western US during severe droughts. A case study from 2020, when D4 drought covered 23% of the western US, aligns with a record 10.1 million acres burned. Time series plots indicate overlapping spikes in drought severity (D3-D4) and wildfire acreage during key periods, such as 2011–2015 and 2017–2021. This trend is particularly evident in California, where persistent drought conditions from 2011 to 2017 closely align with intensified wildfire activity, as well as the record-breaking fires of 2020 and beyond.

From the data, it is evident that there is a strong correlation between drought severity and wildfire activity. As climate change accelerates, addressing this relationship through proactive policies and climate adaptation strategies becomes increasingly urgent. Policy urgency is evident, as regions with frequent D3–D4 droughts, particularly in California and the Southwest, require enhanced fire mitigation strategies and water-resilient infrastructure.

Featured image: CAL FIRE_Official/Flickr.

You might also like: 3 Facts About California’s Climate That Explain the LA Fires

—

References:

– U.S. Drought Monitor
– National Interagency Fire Center
– Williams, A. Park, et al. “Large contribution from anthropogenic warming to an emerging North American megadrought.” Science 368.6488 (2020): 314-318.
– https://en.wikipedia.org/wiki/2011%E2%80%932017_California_drought

For years, media headlines have depicted a dire situation regarding global bee populations. While it is true that honey bee numbers in certain areas, especially in North America and Europe, have plummeted due to habitat loss, pesticide use, and climate change, recent data reveals a more nuanced picture. In fact, bee populations in some Asian countries have been steadily increasing. This contrasting trend prompts important questions: why are bees flourishing in some regions while facing challenges in others?

—

When we examine global trends, we can observe a promising rise in the number of bee stocks. But is it as straightforward as it appears?

A Tale of Two Trends: The Rise and Fall of Bee Populations

By diving deeper and categorizing the data, we can identify conflicting trends. A comparison between North America and Asia highlights these differences.

North America: A Decline in Numbers

In the United States, bee populations have encountered significant obstacles in recent decades. Data shows that honeybee colonies have been on the decline since the 1960s. The primary causes include habitat destruction, pesticide exposure, climate change, diseases, and parasites.

Asia’s Growing Bee Populations: A Different Story

In contrast to the declines observed in the West, some Asian countries have experienced stable or even growing bee populations. Several factors contribute to this positive trend: a diverse natural landscape, a mild climate, a long-standing tradition of beekeeping, and the rise of commercial beekeeping. For instance, China, the world’s largest honey producer, has significantly boosted its managed honeybee populations to satisfy global demand, compensating for the declines in wild bee populations.

Predicting the Future of Bee Populations

The future of global bee populations is uncertain, but there is hope. The following projection outlines potential trends for bee populations over the next decade across various continents.

Our analysis indicates that while managed bee populations in Asia and Africa are likely to grow, North America may face further declines unless significant conservation efforts are implemented.

Key factors influencing these projections include:

• Using probiotics to boost bees’ immunity in the face of rising threats from pests and diseases due to climate change.
• Encouraging a variety of floral environments and protecting natural habitats to provide bees with a steady and plentiful food source.
• Minimizing the use of synthetic pesticides and fertilizers: biopesticides, which come from plants, bacteria, or fungi, can reduce the dangers linked to pesticide application. However, it’s important to assess the safety of biopesticides for bees, as not all are beneficial.
• Promoting sustainable farming methods: Techniques like polyculture, where multiple crops are grown together, and crop rotation, which involves changing crops on the same land over time, can improve biodiversity and strengthen ecosystem resilience.

A Call for a Balanced Perspective

The idea of a global bee population collapse is not entirely accurate. While some regions are experiencing alarming losses, others are seeing increases, demonstrating that proactive measures can lead to positive outcomes. To truly understand these trends, we need a data-driven and region-specific approach rather than relying on generalized predictions of doom and gloom.

If regions facing declines adopt stronger conservation policies and sustainable farming practices, they could stabilize and even boost their bee populations in the years ahead. At the same time, countries with growing bee populations must stay alert to emerging threats to protect their achievements.

Bees play a crucial role in global food security and biodiversity, and their future hinges on our capacity to adapt, innovate, and safeguard their habitats. Instead of concentrating solely on declines, we should also examine and replicate the success stories of thriving bee populations around the globe.

Data for the graphs in this article was retrieved from FAOstat.

This article explores trends in past hurricane data to understand the significance of these warnings and track the evolving nature of hurricanes over the years. As the intensity of hurricanes grows due to the climate crisis, so too does their damage potential and, along with it, the financial costs we deal with in the wake of destruction.

—

Last month, Hurricane Beryl crashed into Texas, taking out power for more than 2 million people. Its sustained high speed winds brought excessive rainfall and flooding inland while creating a deadly storm surge reaching 4-7ft in places along the coast. 36 people have died Texas and Louisiana after 33 deaths across the Caribbean. An initial estimate put the total economic loss from Hurricane Beryl in the United States alone between $28 to 32 billion. This not only includes storm damage but also wage loss, interruptions of supply chains, and flight delays and cancellations.

Hurricane Beryl reached Category 5 status with winds of 165 mph, marking an early start to the hurricane season and becoming the earliest Category 5 hurricane ever recorded. However, it is not necessarily an anomaly. In May, the National Oceanic and Atmospheric Administration (NOAA) issued a warning for an “above average” hurricane season in the North Atlantic, which, according to the World Meteorological Organization, would make 2024 the ninth consecutive year of “above average” hurricane seasons.

This article explores trends in past hurricane data to understand the significance of these warnings and track the evolving nature of hurricanes over the years. As the intensity of hurricanes grows due to the climate crisis, so too does their damage potential and, along with it, the financial costs we deal with in the wake of destruction.

Data & Categorization

The analysis in this article utilizes data from NOAA’s International Best Track Archive for Climate Stewardship (IBTrACS) Project and the Costliest US Tropical Cyclones Listing. The IBTrACS Project compiles comprehensive tropical cyclone data gathered from agencies worldwide, while the listing concentrates on the damage costs incurred in the US due to these cyclones.

This study focuses specifically on hurricanes occurring between 1980 and 2024. Instead of relying on standard classifications, our analysis adopts a modified version of the Saffir-Simpson scale, which includes a newly introduced Category 6, as proposed by recent research. This addition aims to more accurately capture the exponential increases in tropical storm wind speeds, which has been observed more frequently in recent years.

The table below contains a detailed description of the scale that was used.

International Trends

Upon initial observation, the data indicates that the total number of hurricanes is not rising over time. In fact, there appears to be a slight decline in the number of hurricanes, with fewer storms on average each year. The following graph illustrates this trend, showing that the frequency of hurricanes has remained fairly consistent across decades.

Looking at the data across different regions, we can see a similar trend. Most ocean basins, with the exception of the South Atlantic and North Indian, experience around 10 hurricanes per decade. As of May 2024, most basins have recorded approximately 5 hurricanes, seemingly on course to reach the expected 10 hurricanes by the end of the decade.

So this raises the question, what is an above average hurricane season if hurricanes aren’t becoming more frequent?

There has never been a Category 6 hurricane before the year 2013.

When zooming out to observe 15 year intervals, the same pattern unfolds across our international data; there is a nearly identical number of Category 1-4 hurricanes occurring over each time period. However, there is one alarming shift in recent years: the emergence of Category 6 hurricanes. Before 2013, no recorded hurricanes ever met the threshold for Category 6. These hurricanes are characterized by higher wind speeds capable of much more catastrophic damage, especially near coastal areas.

With a focus on the recent emergence of Category 6 hurricanes, our investigation then turned to examining the average wind speeds of hurricanes over time. As depicted in the graph below, there appears to be an ongoing upward trend in average wind speeds by year.

To validate this trend, we conducted further analysis to assess the correlation between average wind speed and year. The calculated correlation coefficient (~0.57) indicates a moderately strong relationship, suggesting that hurricanes have indeed been getting stronger on average in recent years.

According to our linear model, average wind speeds increase by approximately 0.217 knots per year on average. Our statistical tests confirm that this increase is statistically significant (with a p-value of 0), and the narrow confidence interval (0.121-0.314) around the coefficient of 0.217 reinforces the precision of our linear estimate.

These findings underscore the concerning trend of escalating hurricane intensity over the years, potentially leading to more catastrophic damage from these severe storms.

Case Study: U.S.A

Financial impact is one of the metrics used to quantify the resulting damage of tropical cyclones and conceptualize their impact on existing infrastructure. As such, we used the United States as a case study to evaluate the economic impact of hurricanes and identify comparable trends. Our focus was on storms that caused 1 billion dollars of damage or more when they occurred–these costs were adjusted based on the 2024 Consumer Price Index. In total, our analysis encompassed 52 storms occurring between 1980 and May 2024 that had a direct impact on the United States.

The chart below illustrates how the cost of damage caused by hurricanes in the United States has skyrocketed over time. Every 15 years, this amount grew by over 300 billion dollars. In the most recent period alone, costs totaled to around 800 billion dollars. This staggering figure is more than seven times the total costs incurred from 1980 to 1994.

Analyzing the data by decade revealed even more striking disparities. The damage costs incurred in the 2000s exceeded those of the 1980s by more than tenfold, while the 2010s’ costs were nearly thirteen times that amount. In the early 2020s, we have not only surpassed the individual damage costs of the 1980s and 1990s but also their combined total.

Alongside growing costs, the frequency of hurricanes causing over 1 billion dollars of damage has continued to rise as time goes on. When looking at the chart below, we can see that each decade since 2000 has exceeded its 1980s and 1990s counterparts. While we are only halfway through this current decade, the United States has already experienced 15 hurricanes that caused over 1 billion dollars of damage—more than any previous decade. This number is only expected to rise.

Conclusion

Hurricane Beryl marks a violent start as the first major hurricane of 2024. With wind speeds surpassing the Category 6 threshold, it underscores the alarming trends outlined in this study. These trends are a stark reminder that none of our weather phenomena is isolated from the escalating impacts of the climate crisis. Long-term greenhouse gas emissions caused by actions such as burning fossil fuels have significantly warmed the ocean. The higher wind speeds and intensity of hurricanes are largely fueled by rising ocean temperatures, which allows them to gain more speed and power.

As storms like Beryl continue to exhibit increased intensities and destructive potential, reducing greenhouse gas emissions and addressing the root causes of the climate crisis remain imperative to protecting the world from future disasters.

More on the topic: What Are Tropical Cyclones? Hurricanes and Typhoons, And Their Link to Climate Change

The CBAM legislation addresses the European Union’s (EU) concerns about domestic firm competitiveness and carbon leakage. By incentivising EU trade partner firms to reduce emissions and EU trade partner states to adopt carbon-forward policies, the legislation may be instrumental in reducing carbon emissions across the globe to meet the 1.5C target. However, risks of retaliation from the developing world may do more harm than good.

—

What Is the EU’s Carbon Border Adjustment Mechanism?

The European Union’s Emissions Trading Scheme (EU ETS), rolled out in 2005 and gradually expanded since then, sets a cap on greenhouse gases (GHGs) emitted by firms based in the EU by providing them a quantity of Emission Allowances. Firms producing relatively lower emissions can sell their excess allowances to firms producing relatively higher levels of emissions, creating a secondary market for trading these emission allowances.

Announcements geared towards expanding the scope of the EU ETS, along with the phaseout of free allowances, raised concerns that the subsequent increased costs to domestic producers would place them at a competitive disadvantage compared to foreign firms producing goods without a corresponding carbon price. 

Competitive disadvantages for domestic firms could result in carbon leakage, a process in which domestic enforcement of a price on carbon increases emissions elsewhere, either by domestic firms moving production offshore or external firms increasing production to meet the growing demand. In response, the European Commission legislated the Carbon Border Adjustment Mechanism (CBAM) to enforce a carbon price on imported goods in certain industries.

The rationale for the CBAM is that a carbon “tax” on imports of goods covered under the ETS would maintain the domestic competitiveness of European Union (EU) firms and prevent carbon leakage towards EU Trade Partner Countries – countries exporting CBAM related goods to the EU – thereby not undermining the EU’s ambitious emissions reduction goals. 

Initially, the six industries most susceptible to carbon leakage – i.e., emissions-intensive and trade-exposed (EITE) industries – have been included in the mechanism. During the transition phase between October 2023 and late 2025, importing firms will have to report the quantity and emissions intensities of CBAM goods imported. The definitive phase, which is set to commence in January 2026, will require importing firms to pay a price on the carbon emissions on imported goods, linked to the price of EU ETS Emission Allowances.

More on the topic: The Implications of the EU Carbon Border Adjustment Mechanism on the Environment and Global Trade

Firm-Level Incentives and Emissions Reductions

The CBAM Implementing Regulations (2023) require trade partner firms to report their direct as well as indirect emissions, such as those from the generation of electricity consumed for production and from certain production precursors. There are concerns that to minimize their carbon tariffs, firms may engage in “reshuffling”, i.e., exporting to the EU from plants with “cleaner” production (relatively lower direct and indirect emissions) and sending goods from “dirty” plants (relatively higher direct and indirect emissions) to other parts of the world. This could create a two-tier system where relatively “clean” goods are exported to CBAM regimes while “dirty” goods are shipped to non-CBAM regimes with lax climate policies, potentially resulting in relatively lower emissions reductions than anticipated. 

Firm-level emissions reduction incentives also depend on their levels of export dependence on the EU and the carbon emissions intensities of their exports relative to EU goods. The figure below, produced using data from the World Bank CBAM exposure index, provides some useful insights.

Countries with higher levels of export dependence on CBAM products to the EU (measured on the horizontal axis as the share of a country’s total exports of CBAM products that go to the EU) may find it difficult to find alternative export markets. Some African countries (Cameron, Zimbabwe and Mozambique) and Non-EU European countries (Albania, Belarus and the UK) have more than 50% of their CBAM related goods going to the EU. The vertical axis represents CBAM product related emissions intensities relative to EU manufacturers (measured as the excess carbon price to be paid per dollar of CBAM product exports to the EU relative to EU manufacturers). Higher relative emissions intensities signal a competitive disadvantage for a country’s producers and vice-versa. Trinidad and Tobago and Colombia face disproportionately higher (26%) and lower (-11%) relative CBAM tariffs, respectively.

More generally, firms in countries positioned higher up and towards the right would be more susceptible to CBAM tariffs. Countries including India, Georgia, and Ukraine, although not as export dependent, face a cost increase greater than 10% relative to EU manufacturers. Zimbabwe and Belarus face both significant export dependence and higher CBAM tariffs.

Policymaker Responses

Trade partner governments highly dependent on CBAM product exports to the EU may be incentivised to implement their own carbon pricing schemes. In Figure 2, the size of the bubbles indicates the share of a country’s GDP dependent on CBAM product exports to the EU, serving as a proxy for CBAM’s potential spillover effects on a country’s economy. Governments in relatively highly exposed economies may face less domestic resistance towards carbon pricing initiatives, increasing the chances of adoption. Moreover, implementing their own carbon pricing schemes would not only help firms offset CBAM tariffs, but also provide governments with an additional revenue stream. Some countries, including Argentina, China, Chile and South Africa, already have some form of carbon pricing, with many more following suit, including India, Indonesia, Turkey, Brazil, Mozambique and Zimbabwe.

On the other hand, the negative impacts of the CBAM on developing economies, the largest exporters of EITE goods to the EU, may lead to opposition, and potentially retaliatory measures against the policy. Simulations have shown that, in addition to the regulatory burden of CBAM, such countries could face significant welfare losses, exacerbated by lower safety nets. This shift of the burden of fighting climate change towards developing economies disregards their “historical responsibility” for causing the most amount of climate damage. In addition, indirect “sanctions” on (mostly developing) countries violate the principle of “common but differentiated responsibilities and respective capabilities” by coercing them to take action, which most certainly may be beyond their current means. Unsurprisingly, the CBAM has seen negative reactions from many developing economies, particularly from the BASIC group comprising China, India, Brazil and South Africa.

Looking Ahead

The CBAM is just the beginning of a series of climate-related legislations. Similar carbon border tariffs have been proposed by other countries such as the US, Canada, UK, and Australia. As these carbon border legislations increase in scope and geographical coverage, the road will only get tougher for developing countries and for the future of international climate diplomacy. It is too early to say if CBAM is the right way forward. While such policies may be necessary to reach the Paris Agreement goals, the risks of global disorder and the potential unraveling of the global climate agenda may do more harm than good.