Groundwater is earth’s most important freshwater resource. Currently groundwater is abstracted at rates that exceed recharge from precipitation and rivers. This over-abstraction leads to significant groundwater level drops and losses of groundwater from its storage. When groundwater levels drop, groundwater discharges to steams decline (or even completely stop), decreasing streamflow with likely devastating effects on aquatic ecosystems. Environmental flow limits due to groundwater consumption are reached when streamflow drops below an environmentally critical level required to maintain healthy ecosystems (Gleeson and Richter (2017), de Graaf et al in review).

The two MSc projects advertised here will focus on the effect of groundwater pumping on streamflow. The students will use the global-scale surface-water groundwater model PCR-GLOBWB-MODFLOW (de Graaf et al 2017, Sutanudjaja et al 2018) to simulate i.a. groundwater flows and groundwater-surface water interactions under different water management scenarios (project 1) or focusing on uncertainties in model setup and parameterization (project 2). The model we will use is a physically based high-resolution global-scale hydrological model, that runs over the period 1960-2100 at a super-computer. The model is scripted in PCRaster-Python.

In this project the MSc student will formulate ‘management options’, that will be based on a literature study and/or on hypothetical scenarios. The student will run the global-scale model applying this management options starting in 2010 until 2100 and focus the analysis on the impact of groundwater pumping on streamflow. One example of a hypothetical scenario could be: What if we stop pumping at all? A run like this will show us how fast and where the groundwater system will recover (if it recovers), and thus will provide us with a global overview of the detrimental environmental legacy of groundwater use.

In this project the MSc student will search for an interesting regional-scale case-study area and will construct a groundwater model (similar to the global-scale model). A requirement of the regional scale model to be constructed is that it uses for example calibrated sub-surface parameters, more accurate pumping estimates, etc. As such the regional scale model is assumed to be more accurate than the global-scale model. An example of such a case-study is the Republican river basin (ref), located in the High Plains aquifer (USA). The student will run the newly constructed regional-scale groundwater model, and if needed re-run (part of) the global-model. The analysis will focus on comparing model outcomes. First of all, it will provide us with data where measurements are lacking/ do not exist at all (e.g. groundwater discharge is hardly ever measured) and it will show us how good (or how bad) our current global-scale estimates are. Such a model comparison will give us insight what we should improve in the ‘next generation’ of the global groundwater model.

You should pick this topic: if you are interested in groundwater, if you want to learn how to work with a complex hydrological model, if you are not scared of super-computers, python, scripts, linux, and a lot of nerdy stuff.

The knowledge on modelling obtained in the MSc courses. Good English writing skills.

1) Dr. ir. Inge de Graaf (inge.de.graaf@hydro.uni-freiburg.de) 2) Tba