Climate projections that predicted future climate trends are now being put to the ultimate test: comparison with observed data. Three decades of satellite measurements of global sea level change are now available since 1993, provide an exceptional opportunity to assess the validity of these early predictions. These assessments are essential for demonstrating the impact of human activity on the Earth's climate and to strengthen confidence in model-based future climate projections.

Remarkable accuracy of overall projections

Study focused on the IPCC Second Assessment Report (IPCC-SAR) from 1995/1996, specifically on IS92a scenarioThis scenario was chosen because its projections of atmospheric CO2 concentrations match observed values extremely well (427 ppmv vs. 423 ppmv in 2024). The findings show that The IPCC-SAR median projection for sea level rise was remarkably close to the actual development over the next 30 years., with the extent of the rise being underestimated by only about 1 cm by 2023. The observed acceleration of sea level rise, which doubled between 1993 and 2023, closely follows the median projection of the IS92a scenario. This confirms the robustness of early IPCC projections from the mid-1990s.

Variability in projections of individual components

Although the overall projection was accurate, analysis of the contributions of individual components to sea level rise showed variability.

• Thermal expansion of seawater (steric component) and loss of mass of glaciers and ice caps were predicted fairly well, although the steric rise in sea level was slightly overestimated.
• The contributions of the ice sheets (Greenland and Antarctica) have been significantly underestimated or even neglected in early reports; Antarctica was even projected to see a slight increase in mass.
• Land water storage (e.g., groundwater transfer to the ocean) was considered too uncertain and small to model explicitly at the time, but it turned out to be a more significant factor than anticipated.

Key differences and their implications

The main reason for the underestimation was that early IPCC reports treated ice sheets only using mass balance models. For example, the IPCC-SAR report stated that “the dynamic response can be ignored on timescales of decades to centuries.” This weakness led to dynamic mass loss from ice sheets was not sufficiently taken into account. Mass losses from Greenland and Antarctica accounted for approximately 25 % of global sea level rise from 1992–2018/2017. Interestingly, reassessment of thermal expansion partially compensated for low projections of ice sheet contributions, which led to a relatively accurate overall result that could be considered partially fortunate.

Modeling development and confidence in the future

Deficiencies in ice sheet modelling have been evident in later IPCC reports. For example, IPCC-AR4 largely excluded this component, leading to the lowest sea level projections. Subsequently, IPCC-AR5 explicitly included dynamic ice sheet change, increasing the projections. The most recent assessments even acknowledge the “deep uncertainty” associated with ice sheet break-up processes and the possibility of “low probability but high impact scenarios” where sea levels rise much faster. Despite these limitations, particularly with regard to ice sheets and terrestrial water supplies, The early success of the IPCC-SAR projections builds considerable confidence in future climate projections.This historical test confirms the importance of continued and rigorous monitoring of all relevant components of the climate system. JRi

The study is published in the journal Earth's Future .