The Miocene epoch, spanning 23.03-5.33 Ma, was a dynamic climate of sustained, polar amplified warmth. Miocene atmospheric CO 2 concentrations are typically reconstructed between 300 and 600 ppm and were potentially higher during the Miocene Climatic Optimum (16.75-14.5 Ma). With surface temperature reconstructions pointing to substantial midlatitude and polar warmth, it is unclear what processes maintained the much weaker-than-modern equator-to-pole temperature difference. Here, we synthesize several Miocene climate modeling efforts together with available terrestrial and ocean surface temperature reconstructions. We evaluate the range of model-data agreement, highlight robust mechanisms operating across Miocene modeling efforts and regions where differences across experiments result in a large spread in warming responses. Prescribed CO 2 is the primary factor controlling global warming across the ensemble. On average, elements other than CO 2 , such as Miocene paleogeography and ice sheets, raise global mean temperature by ∼2°C, with the spread in warming under a given CO 2 concentration (due to a combination of the spread in imposed boundary conditions and climate feedback strengths) equivalent to ∼1.2 times a CO 2 doubling. This study uses an ensemble of opportunity: models, boundary conditions, and reference data sets represent the state-of-art for the Miocene, but are inhomogeneous and not ideal for a formal intermodel comparison effort. Acknowledging this caveat, this study is nevertheless the first Miocene multi-model, multi-proxy comparison attempted so far. This study serves to take stock of the current progress toward simulating Miocene warmth while isolating remaining challenges that may be well served by community-led efforts to coordinate modeling and data activities within a common analytical framework.Plain Language Summary As human activity continues to increase atmospheric carbon dioxide concentrations, scientists turn to warm intervals in Earth's history to develop insight into the behavior of the climate system under elevated carbon dioxide and temperature. One such interval is the Miocene epoch which has become increasingly relevant as reconstructions of Miocene atmospheric CO 2 concentrations point to values ranging between current concentrations of ∼400 ppm and those projected for the end of this century under Shared Socioeconomic Pathways 3 and 4. In this study, we evaluate the BURLS ET AL.

Sundaland is the currently partially drowned continental landmass that encompasses Borneo, Sumatra, Java, and the Malay Peninsula. It has episodically been reclaimed by the sea during successive Quaternary glaciations, and is commonly thought to be vertically stable. Combining geomorphological observations with numerical simulations of coral reef growth and shallow seismic stratigraphy, we show that the Sunda shelf is subsiding, and that the intermittent regime of transgressions only prevailed over the past 400,000 yr. Prior to that time, Sundaland was permanently exposed. We relate these drowning events to transient dynamic topography in the Indo-Australian subduction zone. Because the Sunda shelf is very shallow, these new data provide important insights into Pleistocene paleogeography, with implications on the interactions between the solid Earth and climate, oceanography, and dispersal of species, including hominids.

The Miocene epoch, spanning 23.03-5.33 Ma, was a dynamic climate of sustained, polar amplified warmth. Miocene atmospheric CO 2 concentrations are typically reconstructed between 300 and 600 ppm and were potentially higher during the Miocene Climatic Optimum (16.75-14.5 Ma). With surface temperature reconstructions pointing to substantial midlatitude and polar warmth, it is unclear what processes maintained the much weaker-than-modern equator-to-pole temperature difference. Here, we synthesize several Miocene climate modeling efforts together with available terrestrial and ocean surface temperature reconstructions. We evaluate the range of model-data agreement, highlight robust mechanisms operating across Miocene modeling efforts and regions where differences across experiments result in a large spread in warming responses. Prescribed CO 2 is the primary factor controlling global warming across the ensemble. On average, elements other than CO 2 , such as Miocene paleogeography and ice sheets, raise global mean temperature by ∼2°C, with the spread in warming under a given CO 2 concentration (due to a combination of the spread in imposed boundary conditions and climate feedback strengths) equivalent to ∼1.2 times a CO 2 doubling. This study uses an ensemble of opportunity: models, boundary conditions, and reference data sets represent the state-of-art for the Miocene, but are inhomogeneous and not ideal for a formal intermodel comparison effort. Acknowledging this caveat, this study is nevertheless the first Miocene multi-model, multi-proxy comparison attempted so far. This study serves to take stock of the current progress toward simulating Miocene warmth while isolating remaining challenges that may be well served by community-led efforts to coordinate modeling and data activities within a common analytical framework.Plain Language Summary As human activity continues to increase atmospheric carbon dioxide concentrations, scientists turn to warm intervals in Earth's history to develop insight into the behavior of the climate system under elevated carbon dioxide and temperature. One such interval is the Miocene epoch which has become increasingly relevant as reconstructions of Miocene atmospheric CO 2 concentrations point to values ranging between current concentrations of ∼400 ppm and those projected for the end of this century under Shared Socioeconomic Pathways 3 and 4. In this study, we evaluate the BURLS ET AL.

Global variations in reef productivity during the Quaternary depend on external parameters that may alter the global chemical balance in the oceans and atmosphere. We designed a numerical model that simulates reef growth, erosion, and sedimentation on coastlines undergoing sea level oscillations, and uplift or subsidence. We further develop a probabilistic evaluation that accounts for variable vertical ground motion, erosion, and foundation morphologies. Absolute sea level change appears primordial, as productivity must have increased by an order of magnitude since the onset of the glacial cycles, ∼2.6 Ma. But most important is relative sea level change, i.e., eustasy modulated by uplift or subsidence, that rejuvenates the accommodation space and exposes pristine domains of the shore to active reefs at each cycle. Integrated over the long‐term, vertical land motion sets the pace of reef growth: productivity in tectonically unstable domains is thus expected to be up to 10 times higher than in stable regions, if any. We quantify the global length of reef coasts and the probability density functions for slopes and uplift rates. Productivity waxes during transgressions to reach 2–8 Gt  CaCO3/yr and wanes during highstands, which may contribute to increase atmospheric pCO2 by several tens of ppm during deglaciations. Over the last 1.5 Ma, reefs precipitated ∼0.8 × 106 Gt  CaCO3 (∼500 × 103 km3), the equivalent of a 1 m‐thick layer spread over the entire surface of the Earth. This production modulates the calcium budget, for it represents some 30% of the modern Ca flux in the ocean.

Many islands of the eastern Indonesian Archipelago exhibit Late Cenozoic sequences of coral reef terraces. In SE Sulawesi, on the Tukang Besi and Buton archipelagos, we identified 23 islands bearing such sequences. Remote sensing imagery and field mapping combined to U/Th and 14 C dating enable to establish a chronologic framework of the reef terrace sequences from Wangi-Wangi, Buton as well as on the neighbouring, smaller islands of Ular, Siumpu and Kadatua. We identified the terraces from the last interglacial maximum (MIS 5e) at elevations lower than 20 m except on W Kadatua where it is raised at 34 ± 5 m. Such elevations yield low to moderate Upper Pleistocene uplift rates (<0.3 mm yr À1). On SE Buton Island, a sequence culminates at 650 m and includes at least 40 undated strandlines. Next to this exceptional sequence, on the Sampolawa Peninsula, 18 strandlines culminate at 430 m. Dated samples at the base of this sequence (<40 m) yield mean Middle Pleistocene uplift rates of 0.14 ± 0.09 mm yr À1. Extrapolation of these uplift rates compared to the geological setting suggests that the sequences of the Sampolawa Peninsula provide a record of sea-level high-stands for the last 3.8 ± 0.6 Ma. The sequences on SE Buton Island therefore constitute the best preserved long-lasting geomorphic record of Plio-Quaternary sea-level stands worldwide.

Abstract. Based on the CMIP5-generation previous IPSL earth system model, we designed a new version, IPSL-CM5A2, aiming at running multi-millennial simulations typical of deep-time paleoclimates studies. Three priorities were followed during the set-up of the model: (1) improving the overall model computing performance, (2) overcoming a persistent cold bias depicted in the previous model generation, and (3) making the model able to handle the specific continental configurations of the geological past. Technical developments have been performed on separate components and on the coupling system to speed up the whole coupled model. These developments include the integration of hybrid parallelization MPI-OpenMP in LMDz atmospheric component, the use of a new library to perform parallel asynchronous input/output by using computing cores as “IO servers”, the use of a parallel coupling library between the ocean and the atmospheric components. The model can now simulate ~100 years per day, opening new possibilities towards the production of multi-millennial simulations with a full earth system model. The tuning strategy employed to overcome a persistent cold bias is detailed. The confrontation of an historical simulation to climatological observations shows overall improved ocean meridional overturning circulation, marine productivity and latitudinal position of zonal wind patterns. We also present the numerous steps required to run the IPSL-CM5A2 for deep-time paleoclimates through a preliminary case-study for the Cretaceous. Namely, a specific work on the ocean model grid was required to run the model for specific continental configurations in which continents are relocated according to past paleogeographic reconstructions. By briefly discussing the spin-up of such a simulation, we elaborate on the requirements and challenges awaiting paleoclimate modelling in the next years, namely finding the best trade-off between the level of description of the processes and the computing cost on supercomputers.

It is widely accepted that sea level changes intermittently inundated the Sunda Shelf throughout the Pleistocene, separating Java, Sumatra and Borneo from the Malay Peninsula and from each other. On this basis, the dynamics of the biodiversity hotspot of Sundaland is consistently regarded as solely contingent on glacial sea level oscillations, with interglacial highstands creating intermittent dispersal barriers between disjunct landmasses. However, recent findings on the geomorphology of the currently submerged Sunda shelf suggest that it subsided during the Pleistocene and that, over the Late Pliocene and Quaternary, is was never submerged prior to Marine Isotope Stage 11 (MIS 11, 400 ka). This would have enabled the dispersal of terrestrial organisms regardless of sea level variations until 400 ka and hampered movements thereafter, at least during interglacial periods. Existing phylogeographic data for terrestrial organisms conform to this scenario: available divergence time estimates reveal an 8‐ to 9‐fold increase in the rate of vicariance between landmasses of Sundaland after 400 ka, corresponding to the onset of episodic flooding of the Sunda shelf. These results highlight how reconsidering the paleogeographic setting of Sundaland challenges understanding the mechanisms generating Southeast Asian biodiversity.

Abstract. Based on the fifth phase of the Coupled Model Intercomparison Project (CMIP5)-generation previous Institut Pierre Simon Laplace (IPSL) Earth system model, we designed a new version, IPSL-CM5A2, aiming at running multi-millennial simulations typical of deep-time paleoclimate studies. Three priorities were followed during the setup of the model: (1) improving the overall model computing performance, (2) overcoming a persistent cold bias depicted in the previous model generation and (3) making the model able to handle the specific continental configurations of the geological past. These developments include the integration of hybrid parallelization Message Passing Interface – Open Multi-Processing (MPI-OpenMP) in the atmospheric model of the Laboratoire de Météorologie Dynamique (LMDZ), the use of a new library to perform parallel asynchronous input/output by using computing cores as “I/O servers” and the use of a parallel coupling library between the ocean and the atmospheric components. The model, which runs with an atmospheric resolution of 3.75∘×1.875∘ and 2 to 0.5∘ in the ocean, can now simulate ∼100 years per day, opening new possibilities towards the production of multi-millennial simulations with a full Earth system model. The tuning strategy employed to overcome a persistent cold bias is detailed. The confrontation of a historical simulation to climatological observations shows overall improved ocean meridional overturning circulation, marine productivity and latitudinal position of zonal wind patterns. We also present the numerous steps required to run IPSL-CM5A2 for deep-time paleoclimates through a preliminary case study for the Cretaceous. Namely, specific work on the ocean model grid was required to run the model for specific continental configurations in which continents are relocated according to past paleogeographic reconstructions. By briefly discussing the spin-up of such a simulation, we elaborate on the requirements and challenges awaiting paleoclimate modeling in the next years, namely finding the best trade-off between the level of description of the processes and the computing cost on supercomputers.