Thesis symposium 2 July 2020

13:20-14:00 Thesis talk Annette van den Engel

The importance of regional ocean modelling for sea-level projections in the North-West European Shelf Region

Regional sea-level projections are of interest because they can strongly deviate from the global mean sea-level change. In a previous study, dynamical downscaling was found to give more realistic simulations in the Northwestern European Shelf (NWES) region than coarse-resolution global climate models (GCMs) of the Coupled Model Intercomparison Project Phase 5 (CMIP5). Dynamical downscaling uses high-resolution regional climate models (RCMs) to refine the GCM results. Here we investigate the performance of dynamically downscaling sea-level projections in the NWES region with the RCM ROMS. The simulations are performed over the period from 1980 to 2098, on a 1/4 ֯ by 1/4 ֯ horizontal grid resolution. ROMS is forced at the lateral boundaries with the CMIP5 GCM HadGEM-ES ocean component. The ROMS surface boundary is forced by the dynamically downscaled HadGEM-ES atmosphere component with the RCM RCA4 from the EURO-CORDEX database. The downscaled sea-level projections are compared to the HadGEM-ES data and the results from [Hermans et al., 2020], where HadGEM-ES was dynamically downscaled with the RCM NEMO (with a horizontal grid resolution of 1/15 ֯ by 1/9 ֯ ).

We find that dynamical downscaling with ROMS leads to a reduced sea-level rise in the North-Sea for the end of the 21st century compared to HadGEM-ES. The results from ROMS on the shelf compare well with the higher-resolution NEMO results. We also find that there is an overestimation of warming, freshening and the annual variability in the simulations when compared to observations and reanalysis data, which is mainly caused by the GCM model used as forcing. Our results show that with using a relatively low resolution RCM large improvements of the GCM sea-level projections are obtained. Allowing dynamical downscaled sea-level projections with a relatively low resolution RCM reduces computational time.

A.D. van den Engel1,2, T.H.J. Hermans2,3, A.B.A. Slangen2, R.S.W. van de Wal1,4
1IMAU, Utrecht University, 2Royal NIOZ, Department of Estuarine and Delta Systems, and Utrecht University, 3Geoscience & Remote Sensing, TU Delft, 4Department of Physical Geography, Utrecht University


14:00-14:40 Thesis talk Tim van den Akker

Modelling a perennial firn aquifer using MODFLOW 6, a case study on the Lomonosovfonna ice cap

Perennial firn aquifers (PFA’s) are water reservoirs in the pore spaces in firn layers on glaciers and ice caps. PFA’s can hold significant volumes of water, slow down sea level rise, affect the mass balance and the dynamics of glaciers and ice caps and contain microbiological life. Research until now has focused on the detection of PFA’s (mountian glaciers, GrIS, Antarctica and Svalbard) and the modelling of the vertical movement of water from the surface to the PFA, referred to as percolation. In this thesis, we developed a lateral model to simulate horizontal flows in a PFA, using the groundwater flow simulator MODFLOW 6. The area that is modelled is the Lomonosovfonna ice field in central Svalbard, for which there is multiple years of measurements available. As input, the model requires depth, density and meltwater data from the Energy Balance Firn Model (EBFM), and a DEM provided by the NPI. The distribution of hydraulic conductivity is used as a tuning parameter to get the modelled water table height close to the observations. Three model experiments were done: a historical run (1957 - 2019), and two future runs for RCP 4.5 and RCP 8.5 (2019 - 2060). The model calculated an average rise of the water table of 10 meters in the historical run, and 12,5 and 20 meters in respectively scenario RCP 4.5 and RCP 8.5. According to the model, the PFA was already present in 1957. The model predicts the water table to first reach the surface in certain spots in 2035 (RCP 4.5) and 2045 (RCP 8.5). The average residence time per model cell drops in the future scenarios, and the average flow direction follows the surface elevation profile. Further research can e.g. be on the development on an above surface routine for this model, or to combine it with a firn model to simulate aquifer-firn interactions.


15:00-15:40 Thesis talk Jan Bouke Pronk

Lake terminating glaciers experience elevated surface velocities in the Himalayan region

Himalayan glaciers melt into the Ganges and Brahmaputra river basins that provide water to over half a billion people. Upstream areas are likely to be affected substantially by climate change, and changes in melt-water supply will locally have tremendous consequences for downstream populations. About 10% of the Himalayan glacier population terminates into pro-glacial lakes and such lake-terminating glaciers are known to be capable of accelerating total mass losses by a well-studied phenomenon called dynamical thinning. However, evidence for dynamical thinning on Himalayan lake-terminating glaciers is sparse and studies available are only local in nature. Here we present a remote sensing-based, 2017-2019 glacier surface velocity dataset covering most of the Central and Eastern Himalayan glaciers larger than 3km2.  We find that centre flow line velocities of lake-terminating glaciers are more than twice as high as land terminating glaciers (18.8 to 8.24 m/year) We partly attribute this difference to dynamical thinning and show that about half of the lake-terminating glacier, of which most clean-ice, show an acceleration towards the glacier terminus. Many of these clean-ice lake terminating glaciers are disproportionately large and drain into the highly melt-water dependent Brahmaputra basins. With continued warming new lake development is likely to happen and will further accelerate future ice mass losses; a scenario not currently considered in regional projections.


15:40-16:20 Thesis talk Hylke Hoogland

Changes in Arctic precipitation and its effects on carbon emissions from thawing permafrost in the 21st century

Context: The rapidly changing climate in the Arctic induces changes in Arctic ecosystems. The thawing of the permafrost may release large quantities of greenhouse gasses that accelerate warming even more. The Pleistocene Park aims to preserve the permafrost by restoring the ecosystem of the Pleistocene: the mammoth steppe. This will change the evaporation of the region, which will in turn change precipitation downwind. As the release of CH4 from permafrost depends on the water table depth, this will also change these emissions.
Aims: In this thesis, I determine the degree of change in precipitation for the Arctic in various future scenarios. In some of these scenarios the tundra is completely replaced by mammoth steppe. An attempt to translate the precipitation changes to CH
4 emission changes is also made.
Methods: I use statistical models based on the current distribution of vegetation and evaporation therefrom to make predictions of evaporation in future scenarios. The evaporated water will be tracked using U-track to determine where it precipitates. Upscaling the equation for CH
4 emission from Olefeldt et al. (2013) will provide insight in the release of CH4 from thawing permafrost.
Results: The statistical models for evaporation deviate from reanalysis data from GLEAM on average with 20.8%. The precipitation change attributable to vegetation change is found to be much smaller than this average deviation in all scenarios (
<5%), and therefore the resulting precipitation estimates are not significant.

Conclusions: Precipitation that evaporated most recently from tundra regions does not fall in regions where the CH
4emission is estimated to be especially sensitive to added moisture. Although no significant estimate of CH4emission change can be made, the increase in CH4emission due to the large scale implementation of the Pleistocene Park method is therefore unlikely to exceed the reduction in emissions achieved otherwise.


16:20-17:00 Thesis talk Evelyn Workman

Investigating the diurnal cycle of the triple oxygen isotope in atmospheric CO2

Currently, the biospheric component of the global carbon dioxide budget is poorly constrained, this biospheric component includes gross primary production (GPP) and respiration. The biosphere is a large carbon sink for anthropogenic CO2 emissions, so it is important that the contribution of this component to the total carbon cycle is better understood. Measurement of the triple oxygen isotope signature ( Δ17 O) of tropospheric CO2 can help to constrain estimates of GPP as this signature is dependent on the amount of CO2 which is equilibrated with leaf water. The first half of this thesis involves characterising the Δ17 O measurement system in order to prepare it for measuring samples of the Δ17 O signature of atmospheric CO2. However due to the Covid-19 crisis, the laboratory was closed just as the system was ready to measure samples. Therefore the second half of the thesis involves using a mixed layer model to simulate the diurnal evolution of Δ17 O in CO2 in Loboos forest. The diurnal Δ17 O(CO2) budget is also investigated, and the individual contributions from entrainment, soil and plants to the total budget are found. The results from the modelled diurnal cycle can be confirmed by measurements of Δ17 O(CO2) at the same location, and they can also help to indicate which measurements would be useful to take in the future.


Start date and time
End date and time