Atmospheric Dynamics

As anyone can see when looking at this website, I am interested in a lot of things. Most of all, learning new stuff. In my view, constantly learning new things is the one ingredient that keeps boredom away. So within the vast field of atmospheric dynamics, I've been moving around between the large scales of general circulation models, and, more recently, the very small scales of tropical convection.

Tropical convection

The tropics are a very cloudy place. And rainy. And sunny, too. And, of course, hot. There's a lot of things happening, and it is generally happening quite fast, and on small scales. Thus, most global climate models can't properly resolve what is happening and have to use approximations and guesses, which are commonly called "parameterizations".

Actually, not just global climate models are having trouble, but even the ones that were developed specifically for resolving convection are struggling. There are still a lot of phenomena which are not correctly (or not at all) simulated in those models. Thus, I am looking at one of the more sophisticated models, the Unified Model, and evaluating its performance in the tropics. More specifically, the region around the Philippines, which is commonly called the Maritime Continent.

A good many things are happening over the Maritime Continent, as for instance self-organisation of single thunderstorms into big convective systems, there are differences over land and over sea, effects due to very steep topography, diurnal cycles of precipitation which the models somehow don't get right, etc. It is important to get these things right in the future, as the tropical regions are the ones which receive the most energy from the sun, and then export that energy in form of temperature, moisture, and winds to the rest of the globe. And therefore the global weather and climate is directly linked to the tropics, meaning that the performance of global forecasts is directly linked to the performance in the tropics. Here's also the link to the other atmospheric research I am conducting: The tropics mark the locations where air, moisture, and lots of chemical tracers leave the troposphere and enter the stratosphere.

Stratospheric dynamics

I am also interested in stratospheric dynamics and stratosphere-troposphere interaction. The large scale meridional overturning circulation known as the Brewer-Dobson Circulation (BDC) is regulating many observed variables, such as temperature, zonal wind, or trace gas distributions.

The upwelling branch cools the tropics diabatically, creating the tropical cold point, the coldest point in the lower stratosphere. This is where water vapor enters the stratosphere, and the resulting freeze drying is responsible for the extremely dry stratosphere. In the downwelling branch, the opposite is true, and descent warms the high latitudes and polar regions to temperatures above radiative equilibrium.

Tracer transport differs from streamlines defined by the meridional circulation, as mixing along constant potential temperature can be important. The most important tracers for my interest are ozone and water vapor. Both are transported by the circulation, but also influence the circulation by acting on the radiative equilibrium temperature (ozone absorbs incoming solar radiation, water vapor is a potent greenhouse gas). Thus, there is intrinsic nonlinearity between dynamics, trace gases, and radiation.

The BDC is mainly driven by planetary scale Rossby waves, which are generated in the troposphere. If the background wind is zonal and not too strong all the way from the surface into the stratosphere, these waves can propagate and finally break high in the stratosphere. The resulting Eliassen-Palm flux allows air parcels to cross surfaces of constant angular momentum, and propagate polewards. Stationary planetary waves of sufficient scale (wave-one and two) are generated by surface topography, and the Tibetan plateau and to somewhat lower degree the Rockies are the most important generators on Earth's surface. There is therefore a north-south asymmetry in wave drive, and in the BDC.

My focus is on the triangle radiation-tracers-dynamics, and I make uses of a hierarchy of idealized general circulation models to investigate sensitivities and cause-and-effect relationships in the stratosphere. An idealized model allows to concentrate on a few parameters, keeping everything else constant. The result is a basic understanding, which starts at the leading-order processes for the simplest models, and is refined with every additional process a given model in a hierarchy can capture.

In this spirit, I have been developing a stratospheric description in a dry model, which which based on offline radiative perturbation calculations. For the next step in the hierarchy I have recently developed a moist model, which adds full radiative effects, but water vapor is the only active tracer (carbon dioxide and ozone are held fixed), and there are no clouds. Thus, processes where the interaction between dynamics and radiation due to water vapor is important can conveniently be studied with such a model. As an example, I am investigating the annual cycle in the upper tropical troposphere, where the dry model predicts an important contribution from extratropical baroclinic activity, and the moist model adds estimates of the relative importance of local water vapor feedbacks in both eddies and the zonal mean.


Scientific work can be difficult to communicate. This is true for communication to a broader public, at it is increasingly the case even within a given scientific community. Specialization of science often results in scientists having difficulty following the work of fellow scientists that are not working in the exact same field.

Reason enough for me (besides the fact that I enjoy it) to produce simplified schematics and animations to explain the science at hand in easy-to-understand visual language. With time, I have developed a certain skill to creating 3D animations, simplified models, and scientific data representation, and included such objects in my presentations.

I see this as part of my work, as it is my belief that scientists who are being paid with public funds have a certain obligation to inform the public and decision takers about the outcome of their research. Now, there are many who are much more skilled at producing scientific data than representing it (and do not enjoy that part of our work), and it is probably best not to force those scientists to spend time on visualization.

But I do enjoy it, and so I took part in Princeton University's Art of Science competition, and won the best picture award in 2013, and was in the official 2014 video selection. I am also giving workshops on scientific visualization, and started programming software to make visualization achievable for other scientists, and made it freely available on GitHub or the publications section.

Plasma Physics

For my PhD was looking at climate change from a different point of view, namely the side of the quest for technological solutions. Doing computational plasma physics, my work was part of the research in thermonuclear fusion devices, such as tokamaks and stellarators. I wrote a particle-in-cell Monte Carlo code, and coupled it to a electromagnetic wave propagation and a magnetohydrodynamic equilibrium code. With this, I was among the first to be able to perform self-consistent ion cyclotron heating simulations in fully three-dimensional plasma devices. The results could directly be applied to stability experiments and the proposition of a tool for increasing fusion performance.