Arctic sea ice occupies a pivotal position in the Earth’s climate machinery. Its bright, reflective surface bounces solar energy back into space, whereas the darker ocean water that becomes exposed when ice melts absorbs that energy instead. This contrast makes sea ice a powerful amplifier of climate change: less ice leads to more absorbed heat, which in turn leads to still less ice. The implications do not stop at the Arctic Circle. Shifts in sea ice extent can influence atmospheric circulation and oceanic patterns far beyond the polar region, which is why scientists, policymakers and industries alike place a premium on anticipating how much ice will cover the Arctic months before it happens.

Forecasting sea ice on seasonal timescales, however, has long been a stubborn problem. The Arctic is changing rapidly, and the processes that govern ice growth and melt span the atmosphere, the ocean and the ice itself, interacting across scales that challenge even sophisticated physics-based simulations. Francesco Paparella, inaugural director of Mubadala ACCESS at NYU Abu Dhabi and senior author of the study, described the challenge directly. “Forecasting Arctic sea ice several months ahead is a difficult problem, and an increasingly important one as the Arctic continues to change,” he said. “Our approach is deliberately simple, but it performs competitively with much more complex forecasting models.”

The core idea behind RAP is rooted in one of the oldest techniques in the forecasting tradition: analog methods. Rather than simulating the physics of the atmosphere, ocean and ice, the algorithm looks backward. It searches the historical record of Arctic sea ice extent for past episodes in which conditions resembled the present state of the system. Once it has identified those analogous situations, it examines what actually happened next in each case and uses those outcomes to construct a set of possible futures. The result is not a single deterministic prediction but an ensemble of forecasts, each representing a plausible trajectory grounded in what the Arctic has actually done before under similar circumstances.

That ensemble structure is what gives RAP its second major advantage: an honest accounting of uncertainty. When the individual forecasts in the ensemble diverge widely, the spread signals that the current situation is genuinely hard to predict, and users can treat the forecast with appropriate caution. When the forecasts cluster tightly, the method is effectively saying that history offers a clearer lesson about what comes next. This built-in uncertainty estimate is rare among simple statistical tools and addresses a persistent weakness in seasonal prediction, where a single number without a confidence range can mislead more than it informs. Paparella emphasized that this transparency is central to the design. “Importantly, it also provides an estimate of its own uncertainty, making it a useful and transparent benchmark for evaluating future forecasting methods,” he said.

The performance evidence comes from direct comparison with the operational forecasting community. The researchers found that RAP produced forecasts with a level of skill comparable to models used by the Sea Ice Prediction Network, the collaborative effort that coordinates seasonal sea ice outlooks from teams around the world. For September, the month when Arctic sea ice reaches its annual minimum and the most closely watched target in seasonal forecasting, RAP’s forecast error was comparable to that of 34 models used for seasonal prediction. Matching the median performance of a few dozen dedicated forecasting systems with a method that relies on nothing more than the historical ice record is a striking result, and it raises pointed questions about how much of the available predictive signal the complex models are actually capturing.

The contrast with physics-based approaches is worth unpacking. Conventional seasonal sea ice models simulate the dynamics of the atmosphere, the ocean and the ice pack, solving equations for momentum, heat and mass transfer across a gridded representation of the Arctic. These simulations demand substantial computational resources, expert maintenance and careful calibration, and their outputs can be difficult to interpret when they go wrong. RAP, by comparison, uses only the historical record of sea ice extent. Its logic can be inspected directly: here are the past situations that resemble today, here is what followed, and here is how spread out those outcomes were. That interpretability, the researchers argue, is not a cosmetic virtue but a scientific one, because it makes the method’s assumptions visible and its failures diagnosable.

This is precisely why the team proposes RAP as a benchmark for the next generation of forecasting systems, including the rapidly proliferating artificial intelligence models now being applied to climate prediction. Machine learning approaches to sea ice forecasting have grown in number and ambition, but evaluating them rigorously requires a baseline that is simple, reproducible and honest about its uncertainty. “The value of RAP is not that it replaces more sophisticated models, but that it gives us a clear standard against which they can be tested,” Paparella said. “If a much more complex model cannot outperform such a simple approach, that tells us something important about how much additional predictive information that complexity is providing.” In other words, any model claiming an edge over RAP must demonstrate that its added machinery translates into measurably better forecasts, not just added sophistication.

The study, published in Scientific Reports under the title describing random analog prediction as a benchmark for seasonal Arctic sea ice extent forecasting, was based on an observational analysis of the sea ice record rather than new field measurements. The work also carries a regional significance that extends beyond the science itself. The United Arab Emirates has been expanding its investment in polar and Arctic research, and Mubadala ACCESS was established as a center dedicated to climate and environmental science in that context. Paparella noted that RAP could support the country’s growing polar research activities by providing a simple, low-cost tool for seasonal sea ice forecasting, one that institutions without access to supercomputing infrastructure could still run and interpret.

The broader lesson of the study resonates across the data-driven sciences. As models grow larger and more intricate, the community needs reference points that are easy to understand and hard to game, and a method that predicts the future by remembering the past provides exactly that. RAP does not claim to capture the physics of a warming Arctic, and its skill will ultimately depend on how the region’s behavior continues to evolve as conditions move beyond the range of the historical record. But as a measuring stick, it sets a bar that any seasonal forecasting system, whether built on fluid dynamics equations or neural networks, must now clear. For a problem as consequential as anticipating the retreat and advance of Arctic sea ice, a transparent tool that matches dozens of complex models while quantifying its own uncertainty is a meaningful addition to the forecasting toolkit, and a reminder that in climate science, elegance and rigor are not opposites.

Subject of Research: Seasonal forecasting of Arctic sea ice extent using an analogue-based statistical algorithm

Article Title: New NYU Abu Dhabi algorithm could predict Arctic sea ice up to nine months ahead, offering new climate insights

Article References: New NYU Abu Dhabi algorithm could predict Arctic sea ice up to nine months ahead, offering new climate insights. (n.d.). Original publication

Image Credits: AI Generated

DOI: Not provided

Keywords: Arctic sea ice, seasonal forecasting, Random Analogue Predictor, NYU Abu Dhabi, climate science, uncertainty quantification, Sea Ice Prediction Network, Scientific Reports, analog forecasting, polar research, climate modeling, benchmark models

Chuyue Xu and Yanling Wu of Hohai University set out to answer a question that operational forecasting centers have largely been unable to tackle: how has the predictability of the boreal summer IOBM changed across different climate regimes? The obstacle has always been data. State-of-the-art dynamical prediction systems, such as the North American Multi-Model Ensemble, typically only produce hindcasts beginning around 1980, because running coupled ocean-atmosphere models backward through decades of history is computationally prohibitive. That short record makes it nearly impossible to detect whether forecast skill itself rises and falls from decade to decade. To get around this bottleneck, the researchers turned to a technique called the Model-based Analog Forecast, or MAF, method, which borrows its power from a library of long climate model simulations rather than from expensive new computations.

The logic behind the MAF method is elegantly simple. It rests on the assumption that sufficiently similar climate states tend to follow similar evolutionary pathways. For any observed initial state of sea surface temperature anomalies across the Indo-Pacific, the method searches a vast library of coupled general circulation model simulations for the twenty closest analogs, and then simply follows what happened next in those simulated worlds. The library in this study was assembled from the last 500 years of pre-industrial control simulations from 29 CMIP6 models, concatenated into a 14,500-year multimodel time series. Because the forecasts unfold entirely within the model’s own phase space, the method sidesteps the initial shock that plagues assimilation-initialized dynamical forecasts, in which observations are abruptly injected into a model and generate spurious transient adjustments.

Sensitivity tests shaped the final design of the forecasting system. The researchers found that matching analogs over the full Indo-Pacific domain, spanning 20 degrees south to 20 degrees north and 40 degrees east to 90 degrees west, produced substantially better and more stable skill than restricting the search to the Indian Ocean alone, a clear indication that Pacific ENSO signals are essential for initializing the Indian Ocean’s subsequent evolution. Ensemble size also mattered: skill improved rapidly as members were added and saturated at twenty analogs, beyond which additional members offered no statistically significant gain. The forecasts were verified against the Hadley Centre’s HadISST observational dataset, with wind and precipitation benchmarks drawn from ERA5, and all fields were interpolated onto a common two-degree grid with linear trends removed, so that the reported skill reflects only internal variability rather than external forcing.

The validation results were striking. Across all calendar months, the MAF method outperformed the majority of NMME dynamical models, and for the boreal summer IOBM index specifically, it achieved anomaly correlation coefficients above 0.6 at all lead times from one to twelve months, comparable to or even exceeding individual state-of-the-art dynamical models. Skill remained highest in the southwestern Indian Ocean, where downwelling Rossby waves deepen the thermocline and preserve persistent subsurface heat anomalies, and in the North Indian Ocean, where strong air-sea coupling anchored by monsoon-related feedbacks extends predictability. With the method’s credibility established, the team could finally ask the question that had been out of reach: does the predictability of the summer IOBM itself change over time?

The answer is an emphatic yes. Using a 21-year sliding window to filter out interannual noise while preserving low-frequency signals, and applying a change-point detection algorithm, the researchers identified statistically significant transitions in forecast skill at 1978 and 2000. Three distinct regimes emerged. Before 1980, useful forecasts extended only about four months ahead. From 1980 to 2000, skill surged: anomaly correlations consistently exceeded 0.6, and normalized errors stayed below one standard deviation of the observed index for lead times up to a full year, with correlations above 0.8 at most lead times. After 2002, skill collapsed back to levels resembling the pre-1980 era. Crucially, a simple persistence forecast, though far less skillful overall, showed the same decadal rise and fall, confirming that the regime shifts reflect genuine changes in the IOBM’s intrinsic predictability rather than artifacts of the analog method.

What drives these swings? The strongest clue came from the strength of the teleconnection between ENSO and the IOBM, measured as the correlation between the summer basin index and the preceding winter’s Niño3.4 index. The correspondence was remarkable: the time series of forecast skill correlated at 0.92 with the strength of this ENSO-IOBM link. During the high-skill epoch of 1980 to 2000, El Niño events were strong enough to leave a deep imprint on the Indian Ocean. The winter El Niño forced anticyclonic wind anomalies over the southeastern Indian Ocean, exciting westward-propagating downwelling Rossby waves that warmed the southwestern Indian Ocean substantially. That warming then activated the wind-evaporation-SST feedback, a self-reinforcing loop in which weakened surface winds suppress evaporative cooling, allowing the basin-wide warming to persist through the summer monsoon onset and into boreal summer, delivering a strong, predictable signal to forecasters.

During the low-skill periods, by contrast, the chain of events broke down early. Weaker ENSO forcing failed to generate significant southwestern Indian Ocean warming or a well-defined wind-evaporation-SST feedback. The characteristic C-shaped wind anomaly over the tropical Indian Ocean never materialized in spring, the basin-wide warming terminated prematurely, and no significant anticyclonic anomaly formed over the Northwest Pacific in the following summer. The recent low-skill period from 2002 to 2022 followed essentially the same script as the 1958 to 1978 era. The researchers also examined alternative explanations, including decadal changes in ENSO intensity, the depth of the southwestern Indian Ocean thermocline, and the amplitude of the IOBM itself. While the high-skill epoch coincided with stronger ENSO events and a shallower thermocline, neither factor correlated significantly with the skill transitions across the full 75-year record, and the Atlantic capacitor effect, in which the North Tropical Atlantic stores and relays ENSO’s influence, proved negligible once the ENSO signal was removed.

The implications reach well beyond the Indian Ocean. Because the summer IOBM anchors atmospheric anomalies that govern monsoon rainfall and heat extremes across densely populated South and East Asia, the finding that its predictability is nonstationary carries a direct warning for forecast users: the current background climate appears to have reverted to a regime in which summer IOBM signals are intrinsically harder to capture at long lead times, and overconfidence in contemporary seasonal forecasts could prove costly. The study also delivers a constructive message for model developers. Accurately representing the ENSO-IOBM teleconnection, particularly the wind-evaporation-SST feedback that sustains Indian Ocean warming into summer, is essential for advancing forecast skill, and climate models that mishandle this interbasin link will systematically misjudge their own reliability. The MAF framework itself, computationally cheap and free of initialization shock, offers forecasting centers a practical complement to dynamical ensembles and a unique window into how predictability has waxed and waned across three-quarters of a century of climate history, though its authors note it has so far been applied only in hindcast mode and not yet implemented for real-time operational forecasting.

Subject of Research: Decadal variability in seasonal predictability of the boreal summer Indian Ocean Basin Mode and its link to the ENSO teleconnection

Article Title: Decadal change in seasonal prediction skills of the Indian Ocean Basin Mode during boreal summer

Article References: Xu, C., & Wu, Y. (2026). Decadal change in seasonal prediction skills of the Indian Ocean Basin Mode during boreal summer. Climate Dynamics, 64(10), Article 441. https://doi.org/10.1007/s00382-026-08397-5

Image Credits: AI Generated

DOI: 10.1007/s00382-026-08397-5

Keywords: Indian Ocean Basin Mode, ENSO, seasonal prediction, predictability, analog forecasting, CMIP6, wind-evaporation-SST feedback, air-sea interaction, Asian summer monsoon, Indian Ocean, climate dynamics, decadal variability