E. Van Campo

Holocene environmental changes in Bangong Co basin (Western Tibet). Part 4: Discussion and conclusions

Abstract A 12.4 m core taken in Lake Bangong provides a continuous Holocene climatic record. We summarize information on changes in stable isotope and radiocarbon balances in the lake, hydrobiology and vegetal cover in the catchment, deduced from detailed analytical results given in the three preceding papers. The Bangong record is then compared with the environmental history of the neighbouring Lake Sumxi also constructed from multidisciplinary analyses. The two records show a major environmental change at ≈ 10-9.5 ka B.P., attributed to a rapid strengthening of the summer monsoonal circulation which led to wet-warm conditions. This event was followed by a long-term trend toward aridity which culminated around 4-3 ka B.P.. In Western Tibet, maximum monsoon rainfall seems to have occurred from ≈ 9.5 to 8.7 ka B.P. and from ≈ 7.2 to 6.3 ka B.P., as two wet pulses separated by a reversal event centered on 8.0–7.7 ka B.P. Our results broadly agree with the records from Lake Seling in Central Tibet, and Lake Qinghai at the plateaus northeastern margin, and with palaeoclimatic studies in western China which document conditions wetter and warmer than those of today during the early-middle Holocene. The environmental fluctuations recorded in western Tibet appear in phase with climatic changes recognized in tropical North Africa, suggesting that the 8.0–7.7 ka B.P. and the 4.0–3.0 ka B.P. dry events were caused by abrupt disequilibrium in the climatic system.

Holocene environmental changes in Bangong Co basin (Western Tibet). Part 2: The pollen record

Abstract A better knowledge of environmental variations in key regions such as the Tibetan plateau, which plays a fundamental role in the Indian monsoon dynamics and thus on the whole Northern Hemisphere atmospheric circulation, is of great importance. We present here a detailed paleoenvironmental study based on pollen analyses of a sediment core from Lake Bangong. Results on the modern pollen rain are also discussed as they give valuable information for the interpretation of the fossil pollen data. The pollen results of the Bangong sequence and their paleoclimatic interpretation are integrated into a chronological framework on the basis of palynostratigraphical methods. Because of dead carbon in the lake system, it has been difficult to establish a reliable chronology of the Bangong sequence. Using a geochemical model for correcting measured 14C ages, the pollen record from Lake Bangong compares well with the record from Lake Sumxi in Western Tibet. Lake Bangong core provides a continuous Holocene pollen record. Before 9600 yr B.P., the dominance of desertic elements in the pollen record reflects an arid environment. At about 9600 yr B.P., an abrupt shift to steppe vegetation suggests the onset of moister conditions, and from ≈9500 to 6200 yr B.P., proxy data indicate conditions warmer and wetter than today. Proxy data also indicate short-term changes during this interval. Climate instability is suggested by the development of a local palustral flora, which peaks at 7700 yr B.P. and is interpreted as a lake regression, in phase with increase in the percentage of littoral and saline water diatom species between 8600 and 7700 yr B.P. The first humid pulse culminating at 9000-8500 yr B.P. is thus followed by drier conditions around 8000-7700 yr B.P. The interval between 7200 and 6500 yr B.P., regarded as optimal for vegetation growth from the pollen record, corresponds to a second humid pulse. A trend towards aridity is observed during the following period with dry events centered around 5500, 3900-3200 and 700 yr B.P.

A data-based re-appraisal of the terrestrial carbon budget at the last glacial maximum

Abstract An estimate of the glacial-interglacial change in terrestrial carbon storage was calculated from CLIMAP reconstructions of the earths surface at the Last Glacial Maximum (LGM). It implies an increase of ≈ 715 Pg C (430–930 Pg C) from the LGM to present, mainly due to the buildup of boreal and temperate forests. This new attempt to reconstruct the LGM ecosystems and carbon content offers an alternative to a previous data-interpretation-based estimate of ≈ 1350 Pg C change. Our estimate is half-way between that 1350 Pg C value and the previous model-based estimate of no change. Furthermore it is consistent with the mean δ13C shift of the ocean. The greatest biomass and soil carbon increaes occurred around 60°N. Our results show that the interpretation of the data is crucial to the estimation of such a budget. It further demonstrates the need for a more complete knowledge of the complexity of the biogeophysical interactions.

The vegetation carbon storage variation in Europe since 6000 BP: reconstruction from pollen

A method is devised to reconstruct the vegetation ecosystems from pollen data. The twelve ecosystems undisturbed by man used by Olson, Watts & Allison (1985) for Europe are refined into twenty-three more homogeneous entities. Each entity is characterized by three typical pollen spectra. A canonical correlation analysis is performed to optimize the climatic signal in the pollen data. A set of 1249 modem spectra and 330 fossil spectra for 6000 BP is available. Each of these pollen spectra is transformed into canonical coordinates and compared to the sixty-nine typical spectra to obtain the best analogue ecosystem. The carbon density provided by Olsen for each of the ecosystems are refined by using a regression model and a net primary productivity model. The area covered by each of the ecosystems and these carbon densities are used to calculate the total carbon storage in the vegetation. In this study, we do not take into account the carbon stored in the soil. It appears that the carbon storage in the phytomass has not changed since 6000 BP, because the quasi-disappearance of the tundra was compensated by the extension of the deciduous forest which contains less carbon than the conifer forests.

Temporal and Spatial Variations of Terrestrial Biomes and Carbon Storage Since 13 000 yr BP in Europe: Reconstruction from Pollen Data and Statistical Models

Statistical models calibrated from field measurement data are used to reconstruct the past terrestrial carbon (C) storage from pollen data for the last 13 000 yr BP in Europe. The pollen-based climatic and biome reconstructions provide the input data for these statistical models, i.e., annual mean temperature, total annual precipitation, annual actual evapotranspiration, annual potential evapotranspiration and biome type with a spatial resolution of 0.5°×0.5° longitude/latitude. Our reconstructions indicate that the last 13 000 yr BP were characterized in Europe by variations of terrestrial biome and net primary productivity (NPP) at various temporal and spatial scales. For the considered region, our results also suggest that changes in climate have significantly altered the distribution of terrestrial biomes and affected the uptake of CO2 for NPP. However, these changes did not translate into significant C storage change in potential terrestrial biosphere during the Holocene. The largest decrease of terrestrial C storage (compared to modern levels) is found during the late-Glacial period mainly due to the persistence of ice sheets and the small extension of forest.

Past and future carbon balance of European ecosystems from pollen data and climatic models simulations

Abstract As climate changes, there is considerable uncertainty whether northern hemisphere ecosystems will act as atmospheric CO2 sinks or sources. Here, we used statistical models calibrated on field measurements, past terrestrial biomes and climates inferred from pollen and future climatic change scenarios simulated by General Circulation Models (GCMs), to investigate the processes controlling past, present and future CO2 fluxes in the European ecosystems. Our results suggest that climatic change can significantly affect spatial and temporal variations of net primary production and soil respiration, and alter the net ecosystem exchange of CO2. Most of the potential terrestrial biomes in Europe will likely change from a net CO2 sink, which provided a negative feedback for atmospheric CO2 during the last 13 000 yr BP, to a net CO2 source, providing a positive feedback following global warming. The results further illustrate that there is no analogue in the recent past (Late Quaternary) for the probable future ecosystem dynamics.

The variation of terrestrial carbon storage at 6000 yr BP in Europe: reconstruction from pollen data using two empirical biosphere models*

Two empirical biosphere models, the Osnabriick Biosphere Model (OBM) and a simple model based on evap- otranspiration (AET model), were used to reconstruct the terrestrial carbon storage from pollen at 6000 yr BP in Europe. The pollen-based climatic reconstructions of Guiot et al. (1993) and the ecosystems reconstruction of Peng et al. (1994) provide the input data of the two models, i.e. annual tempera- ture, annual precipitation, actual evapotranspiration, potential evapotranspiration and ecosystems with a spatial resolution of 0.5? longitude/latitude. For the region considered, our results suggest that there are no significant variations in the terrestrial carbon storage between 6000 yr BP and present due to a compensation of the forest extension towards the north by the replacement of conifer forest (with higher carbon) by deciduous forests to- wards the east. This is consistent with the recent results of Foley (1994) based on climate and biosphere models. The comparison of the two model results with those obtained from the carbon densities of Olson et al. (1985) and Zinke et al. (1986) shows a large similarity for vegetation but a higher value of the OBM for the soil.