Working Groups

WG Marine sediment archives and ocean modeling

This working group investigates long-term and abrupt changes in the African and Asian palaeo-monsoon and its interactions with high-latitude climate variability since the last glacial period. We combine biogeochemical and micropaleontological proxy data from sediment cores of the Arabian Sea, Red Sea and Mediterranean Sea with results from ocean modeling. Biogeochemical data from a well-dated sediment record of the northern Arabian Sea document non-linear interactions between the subtropical westerly jet (STWJ) and the Indian summer monsoon (ISM) during the Holocene (Figure 1). The displayed proxy records suggest a gradual southward displacement of the ISM during the early to middle Holocene. This general shift was punctuated by phases of intensified STWJ, which can be correlated to cold events in the North Atlantic realm. The STWJ influence became dominant between 4600 and 3000 years ago. After this period, the input of terrestrial lithogenic matter into the Arabian Sea was amplified by increased soil erosion as consequence of intensified agricultural activities of the Indus civilization. We presently extend our proxy records into the last glacial period in order to capture the full range of interglacial and glacial climate variability.

Figure 1: Environmental and anthropogenic interactions illustrated by proxy records from the northern Arabian Sea (Burdanowitz et al., 2021 and references cited therein, doi:10.5194/cp-17-1735-2021). (a) Different phases of the Indus civilization and time periods of interpreted natural and anthropogenic modified vegetation (green bar and arrow) in comparison to (b) average chain length (ACL) of the n-alkanes homologues 27–33 (black line, fivepoint running mean) and (c) lithogenic mass accumulation rate (LIT MAR). The ISM humidity index (blue) and CH4 concentration (red) reconstruction based on Greenland ice cores are also depicted in (b). (d) Bond events (numbered gray bars) based on the HSG drift ice record from the North Atlantic.

Sediment proxy records from the Mediterranean Sea show pronounced changes over the last 21,000 years. Local variations in precipitation and river runoff and more remote drivers such as temperature and salinity variations in the North Atlantic entering the Mediterranean at the Strait of Gibraltar shape the basin circulation and the biogeochemical conditions. However, the prevailing paleoclimate conditions cannot be inferred from sediment data alone. Therefore, we develop and apply a regional ocean-biogeochemical model framework for the Mediterranean Sea to unravel the drivers, which create the circulation variability over the last deglaciation (21,000 years to present day) and leave their imprint in sediment records. One main focus is on the mechanisms that lead to development of organic rich sediment layers (Sapropel) during the early Holocene (started ~10,000 years before present). Such layers are very unusual compared to present day sediments where only very little organic material is deposited at the seafloor. Based on our long-term simulations, we can investigate the roles of the African humid period, the variability of the African monsoon, and the abrupt climate variations in the North Atlantic on the Mediterranean biogeochemistry.

Figure 2: Simulated sea surface temperature (°C) of the Mediterranean Sea during the Last Glacial Maximum (21,000 years before present). The boxes and triangles indicate different sets of estimated temperatures from sediment records (Hayes et al., 2005, doi:10.1016/j.quascirev.2004.02.018). Black area shows modern land distribution, grey indicates dry areas during the last glacial maximum due to lower sea level.

WG Environmental Process Modeling

WG Environmental Process Modeling: Research and development of the Working Group aims to interlink large-scale climate modelling with sedimentological analyses, through dynamical downscaling and environmental process modeling. For the mountainous Asian model domain, dynamical downscaling of coarse resolution lateral boundary conditions to convection permitting scales is performed in the framework of CORDEX-FPS CPTP (Convection-Permitting Tibetan Plateau) using the non-hydrostatic mesoscale Weather Research and Forecasting model. Evaluation of high-resolution modeling results against the Himalayan observational network maintained by WG-2 suggests an improved representation of spatiotemporal variations (Böhner et al., 2020). Hydrological and hydrodynamics simulations indicate that the annual water availability from the Hindukush-Karakoram-Himalayan watersheds will increase, but its timings will see distinct changes under various climate change scenarios and under the Paris Agreement 2015 targets, mainly owing to the strengthening of glacier melt regimes (see Figure, Hasson et al., 2019 and reference therein). Under prevailing climatic conditions, sedimentation transport from the Himalayas to the Tarbela Reservoir will decline in the near future (Tarar et al., 2019).

Figure 3: Jhelum Basin: Median and spread of the discharge annual cycles for each hydrological component based on 80 ensemble members obtained from 4 models under 1.5 °C (blue) as well as under 2.0 °C (red) warmer world.

WG Long-term Monsoon Variability

Using Earth system modeling, WP3 tackles the question of monsoon dynamics in the global climate system and the temporal and spatial variability of the monsoon margins.

Response of monsoon margins to abrupt perturbation

Monsoon margins can be modified on very short time scales by volcanic eruptions. Strong volcanic eruptions affect the hydrological cycle, in particular in the tropics, because they affect the energy stored in the atmospheric column and the vertical stratification of moist static energy. In a set of sensitivity experiments under different idealized Pinatubo-like volcanic eruptions in the Max Planck Grand Ensemble framework, we find that the monsoon precipitation declines for at least two summers after the event and that the decline is a function of the emission strength. Generally, volcanic eruptions decrease (increase) tropical (dry-zone) precipitation, resulting in opposite-sign global warming fingerprint (D’Agostino and Timmreck, 2022, in review). The occurrence of multiple volcanic eruptions during the Holocene could be also related to abrupt shifts of monsoon margins, such as during the global-scale aridity event some 4200 years ago.

Figure 4: Composite summer precipitation difference (mm/day) between the simulation EVA40 (tropical volcanic eruption of 30 Tg S) and the simulation EVA0 (the control run without vol-canic eruption). The panels show at the same time the Northern Hemisphere summer (0 - 45◦N, JJA 1992) and Southern Hemisphere summer (0 - 45◦S, DJF 1991/1992) of the first year after the volcanic eruption. Dots represent 95% confidence level. The blue contour is the global monsoon domain in EVA40.

Monsoon margin and the Mediterranean climate

Changes in the intensity of the monsoon circulation can have a remote effect in the extra-tropics, via the Hadley circulation, i.e., the overturning circulation with a rising branch in the tropics and subsidence in the subtropics that supports dry conditions in the hyper-arid, arid and semi-arid regions. The Mediterranean hydroclimate, lying in a transition zone between the wet mid-latitude storm track and the dry subtropical zone, is particularly sensitive to temperature changes and expansion of the subtropical dry-zone and monsoon margins. D’Agostino and Lionello (2020) indentify seasonal changes of Mediterranean hydroclimate in the simulations of the Last Glacial Maximum to future scenarios to be directly linked with changes in the local Hadley circulation. Moreover, D’Agostino et al. (2020) can attribute the persistent drying wintertime conditions in the first 3-decades of the 2100 to a poleward shift of the local Hadley circulation subtropical margin.

Holocene changes in monsoon margins and their impact on global vegetation

Dallmeyer et al. (2021) indentify the monsoon domains and their margins as the region with the strongest Holocene natural vegetation transitions outside the high northern latitudes. Through the impact on the extra-tropical subsidence pattern and the subtropical anticyclones, the changes in the Holocene monsoon intensity also strongly affect the extratropical vegetation distribution. Their study point at a far greater importance of the monsoon systems on the extra-monsoonal vegetation dynamics than previously assumed. Dallmeyer et al. (2021) also show that the global large-scale vegetation patterns change rather linearly during the Holocene. At regional level, however, the simulations also exhibit non-linear, more rapid changes in vegetation. The most striking one is the West African monsoon margin along the southern Sahara experiencing a rapid increase in desert due to rising moisture stress. Not yet known and not yet found in paleo botanic records are rapid shifts in the simulated tree composition that according to the simulations could also occur in South Asia and in the Southern Hemisphere monsoon margins.

Figure 5: Summarizing sketch of simulated main vegetation, climate and circulation chang-es from 8000 years ago to some 2000 years ago. South of 60°N.

In a next step, variations of monsoon margins during the last some 200,000 years will be tackled as outlined in a preparatory study on the African Humid Periods (Duque Villegas et al., 2022, in review).

WG High-resolution simulations

Global long-term simulations using global Earth system models run at a rather coarse spatial resolution and thereby, tend to underestimate the effect of small-scale process and steep orography on the variations of the monsoon margins. The West African monsoon margin is an interesting case. So far, global Earth system models have underestimated northward expansion of the Holocene African Humid Period. In her PhD project, Leonore Jungandreas (2021, Dissertation Universität Hamburg) used the ICON-NWP regional climate model at a spatial resolution as fine as 5 km. She imbedded the ICON model into the global MPI Earth system model used in WP3. She studied the effect of explicitly simulating deep convection on the simulated rainfall patterns in the Holocene Sahel and Sahara using the regional high-resolution ICON model. She not only identified the interaction between soil hydrology and rainfall as key mechanism in understanding the rainfall patterns in the West African monsoon margin. Her simulations also show high rainfall over the steep Tibesti mountain in the midst of the Sahara, thereby helping to solve the riddle of deep crater lakes, which existed in the today arid Tibesti some 9000 to 5500 years ago.