Other meanings of Paleoclimate Modelling Intercomparison Project
Paleoclimatology
The Paleoclimate Modelling Intercomparison Project (PMIP) is a coordinated international effort to compare and evaluate climate model simulations of past climates, spanning from the last millennium to millions of years ago. Established in 1991 under the World Climate Research Programme (WCRP), PMIP provides a framework for testing models against paleoclimate proxy data and understanding the mechanisms of past climate changes. Its results directly inform future climate projections by constraining model sensitivity and feedbacks.1
PMIP originated in 1991 as a parallel activity to the Coupled Model Intercomparison Project (CMIP), focusing on paleoclimate simulations rather than historical or future scenarios.1 The first phase (PMIP1) concentrated on two key periods: the mid-Holocene (6,000 years ago) and the Last Glacial Maximum (21,000 years ago), chosen for their strong climate signals and abundant proxy data. Subsequent phases expanded the scope: PMIP2 (2002–2008) introduced more complex models and the last millennium experiment; PMIP3 (2008–2014) added the last interglacial (127,000 years ago) and the mid-Pliocene (3.3–3.0 million years ago); and PMIP4 (2014–present) integrates with CMIP6, incorporating new experiments such as the deglaciation and the Common Era (past 2,000 years). The project has grown from a handful of modeling centers to over 20 groups worldwide, fostering a collaborative community that bridges modelers and paleoclimate data specialists.
The primary objective of PMIP is to evaluate the ability of climate models to simulate states different from the present, thereby testing their robustness under a wide range of boundary conditions. This is achieved by designing standardized experiments with prescribed forcings—such as orbital parameters, greenhouse gas concentrations, ice sheets, and vegetation—and then comparing model outputs with paleoclimate proxy reconstructions from ice cores, marine sediments, tree rings, and speleothems.2 A key methodological innovation is the use of data–model comparison metrics, such as the skill score, to quantify agreement. PMIP also provides a platform for diagnosing model biases, such as tropical sea surface temperature errors or the amplitude of seasonal cycles, which are often amplified in paleo simulations. By isolating the role of individual forcings, the project helps disentangle natural variability from anthropogenic influences.
PMIP's core experiments target climate states with distinct forcing regimes. The mid-Holocene experiment tests the response to orbital forcing (increased summer insolation in the Northern Hemisphere) and has been used to validate monsoon dynamics and vegetation feedbacks. The Last Glacial Maximum experiment evaluates the model response to large ice sheets, lowered CO₂ (≈190 ppm), and altered aerosols, revealing the role of dust and sea ice. The last millennium experiment (850–1850 CE) examines natural variability driven by solar, volcanic, and land-use changes, providing a baseline for detecting anthropogenic signals. The last interglacial (127,000 years ago) offers a test of warmer-than-present climates with higher sea levels, while the mid-Pliocene (3.3–3.0 Ma) serves as an analogue for future CO₂ levels (≈400 ppm).3 More recent experiments include the deglaciation (21,000–9,000 years ago) and the Common Era (1–2000 CE), which incorporate transient forcing and ice-sheet dynamics.
Beyond headline experiments, PMIP has supported niche investigations such as the response of the Indian monsoon to orbital precession, the simulation of the Younger Dryas abrupt cooling event, and the role of tropical peatlands in the carbon cycle during the Holocene. A lesser-known contribution is the development of paleoclimate data assimilation methodologies, which combine model outputs with proxy records to produce spatially complete reconstructions of past climate fields.4 PMIP also provides the PMIP4 database of more than 100 simulation outputs, freely available for use by the broader scientific community. The project has been instrumental in identifying emergent constraints on climate sensitivity—for example, linking the simulated strength of the Saharan greening in the mid-Holocene to the equilibrium climate sensitivity. Additionally, PMIP experiments have revealed the importance of vegetation–climate feedbacks, which are often underestimated in simpler models, and have highlighted the need for interactive ice sheets in long-timescale simulations.
PMIP has directly influenced the Intergovernmental Panel on Climate Change (IPCC) assessments by providing a paleoclimate perspective on model performance and climate sensitivity. In the IPCC Sixth Assessment Report, PMIP simulations were used to constrain the likely range of equilibrium climate sensitivity (2.5–4.0 °C) and to assess the ability of models to reproduce past warm periods.3 The project has also advanced understanding of the Atlantic Meridional Overturning Circulation (AMOC) stability during the last deglaciation and the mechanisms of abrupt climate change seen in ice-core records. By bridging the gap between proxy-based reconstructions and model simulations, PMIP has fostered a more integrated view of the Earth system, demonstrating that models that fail to simulate past climates are unlikely to produce reliable future projections. The ongoing PMIP4–CMIP6 collaboration continues to push the boundaries of spatial resolution and Earth system complexity.
Help improve the encyclopedia. Reports go straight to the site manager.