The world is grappling with a climate challenge that has defied predictions

It is time to reconsider the massive resources invested in future projections, and focus on locally relevant predictions.

Because of heat wave conditions, fire brigade personnel spray water onto devotees thronging the Shree Jagannatha temple during the Ratha Yatra festival in Puri in June.  | Photo Credit: BISWARANJAN ROUT

As 2023 inches towards its end, literally every day brings another headline on how many records are being shattered. Surprisingly, there are many headlines about the stunning level of global warming, catching even climate scientists unawares. One headline said scientists were amazed by the “gobsmackingly bananas” September heat, and another one called the sudden spike in warmth mysterious. The warming also, of course, served up another opportunity for some climate scientists to say, “We told you so.”

The general public however has been left wondering why this was not only not foreseen but is instead being claimed as a surprise and a mystery. What does it mean for the reliability of future projections of global warming?

Even though this year was not predicted to attain this level of warming, it is already being claimed that we are likely to cross the 1.5°C global warming threshold this year. And yet, it is unclear how long we will stay above this threshold or what the impacts will be. Or even what the next year will bring.

Climate variability and predictions

Year-to-year climate variabilities fall under the realm of climate variability and predictions while the climate Assessment Reports put out by the Intergovernmental Panel on Climate Change (IPCC) fall under climate trends and projections. Neither the projections nor the predictions portended the extent of wildfires, floods, droughts, and so on, that have occurred this year. This year has made it clear that global warming and year-to-year variabilities or natural variability can combine to shock us with sudden warming. The key question is whether the magnitude of the warming is because of the internal heat being moved around or if more energy has come from outside the earth system. 

A view of the wildfire that scorched the town of Tara, Queensland, Australia, in October. | Photo Credit: Hiro Komae/Queensland Fire And Emergency Services/AP

The only heat that can be released within the earth system is from the oceans. No solar activity has been observed indicating any additional energy arriving from the sun. But a cleaner atmosphere can trap more of the sun’s energy. While a cleaner atmosphere is claimed to have contributed to this year’s warming, as of now, we are clueless about the exact reasons for the level of warming seen so far.

There has also been a slowdown in warming recently, from 1998 to about 2012. The global warming rate slowed following the strongest El Niño of the 20th century, and the slowdown lasted for over a decade. This raises the obvious question, what will next year or the coming years look like?

The record-breaking warming is an excellent opportunity to understand what future climate projections are about and how they are different from climate predictions. It is also another warning that much more attention needs to be paid to the so-called “societally relevant” timescales of multiple years for climate predictions. Long-term projections to year 2100 and beyond get much publicity, but their usefulness has to be evaluated critically in light of the warming in 2023 and the lack of information on managing disasters effectively, especially at local scales where the impacts of climate are maximum.

The World Climate Research Programme (WCRP) under the World Meteorological Organization issues annual to decadal climate updates, which do not get as much attention as the IPCC assessment reports. However, the recently released decadal update also failed to capture the magnitude of the warming seen this year.

Numerical or computer models

Numerical or computer models used to predict and project weather and climate are all essentially similar to each other. The models are impressive in their ability to simulate the weather and climate: winds, ocean currents, clouds, precipitation, monsoon, El Niño, La Niña, and so on. And they do this just given the energy coming from the sun at the top of the atmosphere. Models are built to balance the incoming solar energy with the earth’s outgoing thermal energy.

Models are never perfect because they are trying to simulate an infinitely complicated system. The model monsoon tends to be weaker and the model El Niños tend to be weaker and much more regular in time than in nature. Nature gives us an El Niño every two to seven years, but models tend to do an El Niño about every four years. Models have not been able to tell us confidently how the monsoon or the El Niños themselves will change because of global warming.

Boats stranded on the dry bed of the Puraquequara lake in Manaus, Amazonas State, Brazil, on October 6. Brazil was one of the countries in South America affected by the heat wave that struck the continent between July and September. | Photo Credit: MICHAEL DANTAS/AFP

The models are, however, useful for weather to climate predictions and climate projections. We need to understand a few basics of how weather and climate predictions differ from each other before we learn how climate projections are distinct from climate predictions. It is worth remembering the old adage that all models are wrong, but some are useful.

We expect November to be cooler than May because climate cools during the autumn. But a given day in May, say May 5, might turn out to be cooler than November 5 because the weather can deliver surprises. This has led to the saying that climate is what we expect, and weather is what we get. In terms of the dynamics, the atmosphere is much lighter than the ocean, and the heat capacity of air is very low compared with that of water. On a timescale of a few days, the atmosphere can warm up and cool down rapidly, but the ocean warms up slowly and cools down slowly.

Highlights

• 2023 saw record levels of warming across the globe that were not predicted and appear to have caught even climate scientists unawares.
• It is being claimed that we are likely to cross the 1.5°C global warming threshold this year but there is no clarity on how long we will stay above this threshold or what the impacts will be.
• This is an excellent opportunity to understand what future climate projections are about and how they differ from climate predictions and to evaluate their usefulness in light of the warming in 2023 and the lack of information on managing disasters effectively, especially at local scales where the impacts of climate are maximum.

The butterfly effect

These differences are used to simulate and predict the weather with just an atmospheric model since ocean temperatures do not change much in a few days. The weather becomes chaotic after a few days since the atmosphere can go crazy after a few days. This is often referred to as the butterfly effect: the flap of a butterfly’s wings in the Amazon can cause a cyclone in the Indian Ocean, metaphorically speaking, of course. This is just to indicate that weather prediction is limited to a few days.

The atmosphere, however, cannot run away on its own for longer than a week or two since the ocean begins to pull the bridle. The coupled interactions and feedback between the ocean and the atmosphere are what we call climate, the co-evolution of the ocean and the atmosphere. Just as the weather is predictable only for a few days, the climate is mostly predictable for a season or two. Seasonal features such as the monsoon have some predictability while climate modes such as El Niño can be predicted several seasons ahead. For example, the current El Niño was predicted as early as February. This El Niño will peak in December-January-February and is predicted to decay into the spring of 2024.

Macumba beach, in the west zone of Rio de Janeiro, Brazil, on September 24, during the heat wave.   | Photo Credit: Tercio Teixeira/AFP

Some level of skill exists for climate predictions at multiple years to a decade, and this is what motivates the WCRP efforts on its annual to decadal updates. There is much discussion in the scientific literature on heat and precipitation extremes, cyclones, sea level rise, and such by 2100. But this year is a stark reminder of the urgent need for predictions for multiple years since these are the timescales of decision horizons for governments and the private sector.

The goal of this article is not to throw the baby out with the bathwater. Climate projections have their value in alerting governments and societies into action on climate mitigation that defines safe trajectories to navigate economic development and avoid potentially unmanageable outcomes in terms of climate impacts and irreversible outcomes.

What then are the differences between climate predictions and projections? Considering that the evolution of climate beyond a decade is heavily dependent on greenhouse gas emissions, we need to provide some guidance to the climate models on how future emissions may evolve. Socio-economic models are used to guesstimate future emissions on the basis of various assumptions about technologies, populations, and choices by governments and societies in terms of energy sources (fossil fuels vs renewables), climate policies, and even diets. These choices are categorised into the so-called Shared Socioeconomic Pathways, which also determine the warming levels likely to be attained by 2100.

The biggest challenge for climate projections is that the models are quite skilful in producing the warming expected from the increase in greenhouse gas emissions. Model accuracies are assessed by comparing their simulations for the historic period starting from the beginning of the Industrial Revolution to the present. The comparisons of global mean temperatures for the period of 1850-2020 as recorded in the instrumental records and as simulated by the models indicate exactly why we may have surprises such as 2023.

While climate predictions must indicate what will happen next season or next year, climate projections can only depict how warm or cool the global mean temperature may be in the coming years. In other words, the projections simulate only the warming trends and do not capture accurately year-to-year variabilities. This year was supposed to have enhanced warming since an El Niño releases heat from the ocean and adds to global warming. But climate projections mostly do not capture such jumps in warming accurately. This year’s warming is also way beyond what can be expected just from a superposition of global warming and El Niño.

Why has 2023 been so warm?

Why has 2023 been so warm? A few factors may have contributed to this unexpected warming: The early and strong evolution of this year’s El Niño is a major one. Two other candidates are suspected to have contributed to the unexpected spike. The 2022 eruption of the submarine Hunga Tonga-Hunga Ha’apai volcano in the South Pacific Ocean pumped a massive amount of water vapour into the stratosphere, the upper atmosphere that can trap the earth’s outgoing thermal energy. Water vapour is a strong greenhouse gas.

Reduced sulphur dioxide levels in the atmosphere are also argued to have allowed more solar energy to have entered the atmosphere to enhance warming. Sulphate aerosols scatter sunlight and have a cooling effect on the global mean temperature. It is not clear whether all the wildfires also contributed to the warming.

A sign outside the international airport in Phoenix, Arizona, displays an unofficial temperature (42°C) on July 12. Arizona’s Maricopa County, the hottest metropolitan area in the US and home to Phoenix, set a new record on October 19 for annual heat-associated deaths. | Photo Credit: Matt York/AP

The final word on the details is yet to emerge. This underscores another critical detail about climate projections. They are dominated by natural climate variability in the first decade or two into the future, that is to about 2040. Beyond a couple of decades, model uncertainties begin to impact the projections for a few decades, say to about 2060. Beyond this, the projection uncertainties are completely dominated by the scenarios themselves. The COVID-19 pandemic and the ongoing geopolitical perturbations continue to cascade through economies, energy markets, and climate actions, pointing to the great challenges of designing future scenarios and their impacts on emissions and global warming.

That brings us back to wondering about the cost and benefit of investing massive resources in making future projections while we are unsure of even what the next year will bring. How should each country and its states, towns and villages, and businesses prepare for next year or the coming few years? What is the value of projections beyond a few decades and what is the optimum number of years for projections? What decisions are really being taken based on the projections? The biggest question is of course about the scales at which projections offer usable information. Will they ever offer local-scale climate risk information to guide adaptation strategies and financial needs?

The IPCC and all the signatories of Paris Agreement must use the excessive warming in 2023 to seek answers to these existential questions in order to design the future of climate projections. Considering the limited resources and the constant tug of war between economic development ambitions and climate actions, continuing blindly down the same path of climate projections may not be a wise or realistic option.

Raghu Murtugudde is Professor, IIT Bombay, and Emeritus Professor, University of Maryland, US.