Description: Afforestation of the Sahara has been proposed as a climate engineering method to sequester a substantial amount of carbon dioxide, potentially effective to mitigate climate change. Earlier studies predicted changes in the atmospheric circulation system. These atmospheric feedbacks raise questions about the self-sustainability of such an intervention, but have not been investigated in detail. Here, we investigate changes in precipitation and circulation in response to Saharan large-scale afforestation and irrigation with NCAR’s CESM-WACCM Earth system model. Our model results show a Saharan temperature reduction by 6 K and weak precipitation enhancement by 267 mm/year over the Sahara. Only 26% of the evapotranspirated water re-precipitates over the Saharan Desert, considerably large amounts are advected southward to the Sahel zone and enhance the West African monsoon (WAM). Different processes cause circulation and precipitation changes over North Africa. The increase in atmospheric moisture leads to radiative cooling above the Sahara and increased high-level cloud coverage as well as atmospheric warming above the Sahel zone. Both lead to a circulation anomaly with descending air over the Sahara and ascending air over the Sahel zone. Together with changes in the meridional temperature gradient, this results in a southward shift of the inner-tropical front. The strengthening of the Tropical easterly jet and the northward displacement of the African easterly jet is associated with a northward displacement and strengthening of the WAM precipitation. Our results suggest complex atmospheric circulation feedbacks, which reduce the precipitation potential over an afforested Sahara and enhance WAM precipitation.

Description: Observations indicate an expansion of oxygen minimum zones (OMZs) over the past 50 years, likely related to ongoing deoxygenation caused by reduced solubility, changes in stratification and circulation, and a potential acceleration of organic matter turnover in a warming climate. Higher temperatures also lead to enhanced weathering on land, which, in turn, increase the phosphorus and alkalinity flux into the ocean. The overall area of ocean sediments that are in direct contact with low oxygen bottom waters also increases with expanding OMZs. This leads to an additional release of phosphorus from ocean sediments and therefore raises the ocean's phosphorus inventory even further. Higher availability in phosphorus enhances biological production, remineralisation and oxygen consumption, and might therefore lead to further expansions of OMZs, representing a positive feedback. A negative feedback arises from the enhanced productivity-induced drawdown of carbon and also increased uptake of CO2 due to increased alkalinity, which, in turn, got there through weathering. This feedback leads to a decrease in atmospheric CO2 and weathering rates. Here we quantify these two competing feedbacks on millennial timescales for a high CO2 emission scenario. Using the UVic Earth System Climate Model of intermediate complexity, our model results suggest that the positive benthic phosphorus release feedback has only a minor impact on the size of OMZs in the next 1000 years, although previous studies assume that the phosphorus release feedback was the main factor for anoxic conditions during Cretaceous period. The increase in the marine phosphorus inventory under assumed business-as-usual global warming conditions originates, on millennial timescales, almost exclusively from the input via terrestrial weathering and causes a 4 to 5-fold expansion of the suboxic water volume in the model.

Description: Feedbacks determine the sensitivity of the Earth system to perturbations. In the last centuries, the Earth system has been undergoing substantial changes caused to a great extent from the human activity. Nowadays, it is well documented that global warming is mainly driven by the anthropogenic emissions of greenhouse gases. In the future, these emissions will likely continue and lead to further climate warming. Large-scale climate engineering projects are human actions proposed to counteract climate warming. However, these projects could also cause unintentionally perturbations to the Earth system or its subsystems. Scientists apply models to estimate the impact of these anthropogenic forcings on the Earth system in future projections. Therefore it is important that feedbacks are well represented in these models to make reliable predictions. The aim of this PhD thesis is to advance the research in feedbacks and their response to anthropogenic forcing in the Earth system or its subsystems. Two case studies are carried out with a focus on a) the atmospheric feedbacks in North Africa driven by an irrigated afforested Sahara, and b) the biogeochemical feedbacks related to the phosphorus cycle and its link to oceanic deoxygenation. This PhD thesis advances the understanding of the North African climate system and their feedbacks under an artificial large-scale afforestation scenario. It reveals the potential and unintentional side effects of such a climate engineering project. Furthermore, this thesis discusses to what extent human activity could drive the global ocean suboxic or anoxic and reassessed the relevance of the different feedbacks for the deoxygenation of the ocean.

Description: Periodic changes in sediment composition are usually ascribed to insolation forcing controlled by Earth’s orbital parameters. During the Cretaceous Thermal Maximum at 97–91 Myr ago (Ma), a 37–50-kyr-long cycle that is generally believed to reflect obliquity forcing dominates the sediment record. Here, we use a numerical ocean model to show that a cycle of this length can be generated by marine biogeochemical processes without applying orbital forcing. According to our model, the restricted proto-North Atlantic and Tethys basins were poorly ventilated and oscillated between iron-rich and sulfidic (euxinic) states. The Panthalassa Basin was fertilized by dissolved iron originating from the proto-North Atlantic. Hence, it was less oxygenated while the proto-North Atlantic was in an iron-rich state and better oxygenated during euxinic periods in the proto-North Atlantic. This redox see-saw was strong enough to create significant changes in atmospheric pCO2. We conclude that most of the variability in the mid-Cretaceous ocean–atmosphere system can be ascribed to the internal redox see-saw and its response to external orbital forcing.

Description: Previous studies have suggested that enhanced weathering and benthic phosphorus (P) fluxes, triggered by climate warming, can increase the oceanic P inventory on millennial timescales, promoting ocean productivity and deoxygenation. In this study, we assessed the major uncertainties in projected P inventories and their imprint on ocean deoxygenation using an Earth system model of intermediate complexity for the same business-as-usual carbon dioxide (CO2) emission scenario until the year 2300 and subsequent linear decline to zero emissions until the year 3000. Our set of model experiments under the same climate scenarios but differing in their biogeochemical P parameterizations suggest a large spread in the simulated oceanic P inventory due to uncertainties in (1) assumptions for weathering parameters, (2) the representation of bathymetry on slopes and shelves in the model bathymetry, (3) the parametrization of benthic P fluxes and (4) the representation of sediment P inventories. Considering the weathering parameters closest to the present day, a limited P reservoir and prescribed anthropogenic P fluxes, we find a +30 % increase in the total global ocean P inventory by the year 5000 relative to pre-industrial levels, caused by global warming. Weathering, benthic and anthropogenic fluxes of P contributed +25 %, +3 % and +2 %, respectively. The total range of oceanic P inventory changes across all model simulations varied between +2 % and +60 %. Suboxic volumes were up to 5 times larger than in a model simulation with a constant oceanic P inventory. Considerably large amounts of the additional P left the ocean surface unused by phytoplankton via physical transport processes as preformed P. In the model, nitrogen fixation was not able to adjust the oceanic nitrogen inventory to the increasing P levels or to compensate for the nitrogen loss due to increased denitrification. This is because low temperatures and iron limitation inhibited the uptake of the extra P and growth by nitrogen fixers in polar and lower-latitude regions. We suggest that uncertainties in P weathering, nitrogen fixation and benthic P feedbacks need to be reduced to achieve more reliable projections of oceanic deoxygenation on millennial timescales.

Description: Many recent ocean modelling studies have demonstrated the added value of enhanced horizontal resolution, although it comes at a high computational cost. However, few modeling studies of ocean-based CDR have been done at high resolution. Here we assess the effects of model resolution on two simulated ocean-based CDR methods, unequilibrated ocean alkalinity enhancement (OAE) and the direct marine capture (DMC) of CO2 from seawater (with assumed permanent storage), in experiments with the FOCI Earth system model. To do this we utilized two FOCI configurations, one with a 1/2° ocean resolution and the other with a 1/10° ocean nest in the N. Atlantic. Both configurations were run in a series of “paired” experiments with identical climate forcing and CDR deployments. We show that model resolution does appear to matter when simulating OAE and DMC. For OAE, parameterization of physical processes in the coarse resolution version of the model appears to overestimate how long alkalized waters stay in contact with the atmosphere and where they are transported. This results in large differences in OAE efficacy with almost twice as much carbon sequestered when the model resolution is coarse. For the DMC simulations, at one site there were clear differences in the compensating CO2 flux induced by DIC removal, which was again higher with a coarse resolution, while at the other site variability was high and differences were difficult to determine. At both DMC sites there were clear differences in circulation with the two model resolutions, and thus on downstream biogeochemistry. We suggest that well resolving ocean physics may be necessary to best calculate unequilibrated OAE and DMC efficacies and side effects. These results should be confirmed using other models and with different resolutions.