 Okay. So, hello everyone. I'm Katarina Kosovel. I'm a PhD student of Nadia Lkazhagar at the University of Ljubljana and I will present a small part of my research that was done in collaboration with Franco-Multenian Fred Kuczarski. The title is Propagation of Tropical Hitting perturbations to the mid-latitudes and the role of photography. I will start shortly with motivation. Then I will describe the methodology and the results will be presented in two parts. So, for the short-term response and then for the medium-term response, I will briefly look at the role of photography and then summarize. The motivation for my studies is that tropical processes strongly influence mid-latitude weather processes in the medium range and longer forecasts and at the same time analysis uncertainties and short-range forecast errors are largest in the tropics. They are shown to be associated with both balanced and unbalanced circulation across all spatial scales. Little is known about the impact of the tropical analysis uncertainties on the growth of forecast errors in the mid-latitudes. The perturbations in diabetic hitting in the tropics have long-term impact on the forecast of the mid-latitude circulation. So, we aim to quantify the role of unbalanced component of the large-scale circulation response to these perturbations in the tropics and their propagation outside of the tropics. For the experiment, we ran a speedy model and we did an ensemble. Here is the sketch how we did it. So, we had a continuous ampertert run for 32 years and every 1st January for the third year on, the model was restarted with continuous additional heating perturbation and then run for three months. So, we got 30 ensemble members in northern hemisphere winter. This plot shows how the heating perturbation at 500 hPa looks like. It has random component in it, so this is averaged over 90 days. Heating perturbation was adopted from unpublished research of Franco-Multanian Fred Kucharski. Daily mean outputs were analysed with modes decomposition. Modes decomposition works through the projection of fields in physical space to the model space and the circulation is split in balanced and unbalanced components and defined in terms of Eigen solutions of the linearized primitive equations. The balanced part consists of quasi-geostropic cross- bewaves and the unbalanced part projects on to the inertial gravity Eigen solutions that propagate either westward or eastward. We don't need to go that much into details about inertia gravity waves for my research, so we will just say that inertia gravity modes are actually unbalanced modes. The composition software was developed at the University of Ljubljana. More info on it you can get on this web page or come and ask me after. On the web page you can also get the software and browse through the decomposition products of the ECMWF forecasts, such as maps of balanced and unbalanced circulations in tropics and extra tropics, and you can also look at this kind of global energy spectra. This is just an example. How this decomposition works. This is an example. This is one random example filled from our experiment showing geopotential height perturbations and winds, and we can split this field to balanced and unbalanced part. Balanced part is stronger, unbalanced part is weaker, but unbalanced part is important in tropics. Advantage of this decomposition is that you can easily get easy direct access to the unbalanced part. Both can be further split into the particular modes. Balanced part is actually a sum of many Rosby waves. We have three examples here. The Rosby N-N-Wave is more or less confined to tropics, but the others have minimum and maximum in mid-latitudes, and what is important here is that odd modes are symmetric regarding equator and even modes are anti-symmetric. In the unbalanced part the most energetic mode is Kelvin mode and the remaining unbalanced mode are just weaker and I won't go more into details with them. Now we should focus on our experiment. In the first few days the response is mostly to the tropical heating is mostly confined to tropics, and then after the first few days the Rosby wave train forms in mid-latitudes and the perturbations start to propagate around the mid-latitude in a circle. So first I will show an example for short-term response. The total response has been substantially studied in the past. So what we see here is a barotropic beginning of the Rosby wave train that is entering to mid-latitudes and then we have a baroclinic response in tropics that is mainly in the source region and it looks like it consists of Kelvin wave and Rosby N-1. But for confirmation we can look at the composition. The Rosby wave train goes to the balanced part and what looked like a Rosby N-1 wave is also here in the balanced part. We can look at this Rosby N mode separately and we see that the strongest minima and also the other maxima smaller are approximately collocated with minima and maxima in the total balance in the whole balanced response, but they are modified with higher modes. What looked like a Kelvin wave in the total field is actually partly balanced and partly unbalanced, but the unbalanced part itself is to a large extent Kelvin mode. So now to the medium-term response. In the medium-term response this is example for day 14 and just note that the scale for geopotential height perturbations is not the same anymore. The strongest response is in northern hemisphere Rosby wave train, but there is also a weaker Rosby wave train in the southern hemisphere. As the those Rosby wave trains project to the balanced part, balanced part is actually pretty similar to the total except for the tropics and as we can already guess the remaining part is in unbalanced. Looking back to the balanced part, we see that we have here a pattern that resembles again to this Rosby N1 mode and we see that this minima is actually at least partially but to quite some extent Rosby N1 mode. And if I put here the Kelvin mode which is the second important mode in the tropics, we see that even though we are now in day 14, they're still confined to the close to the region, the source region where we have this heating. And if we compare the location to day 3, we see that they practically didn't move out. Actually, we didn't look much on this but we suspect that it's because of the dipolar structure of the heating that we used. Okay. As the tropical response is more or less confined to the source region in the northern hemisphere, in mid-latitudes in general the perturbations do propagate eastward. And as we know in northern hemisphere, the general circulation in the mid-latitudes is very much influenced by the orography, especially the jets. So the logical question was how this orography influenced the path of the perturbations. And we did three additional experiments where the only difference between the main experiments, now the reference experiment and those three, were that part of the northern hemisphere orography was just lower to the maximum 100 meters. The first one was without Himalaya and Tibet. The second one was without Rakis and Greenland. And the third one was the combinations of both. So somehow expected result was that in mid-latitudes the intensity and propagation of Rosby wave train are affected by the large scourography. But rather than showing these plots, I will show something else. As modes decomposition is actually linear and orthogonal, we can do statistics in modal space and get some additional information on what's going on. So these four plots show the variance in a spectral space. This one is the reference with orography and then we have another three. Meridional modes, those numbers just represent the Rosby waves I showed at the beginning. This is all for balanced modes because in mid-latitudes the Rosby wave train, as I showed, projects to the balanced part. And zonal wave numbers just tell us how many waves we have around the globe in zonal direction. So we are interested in mid-latitudes. So we are interested on the modes with the numbers two and more because N1 is in tropics. What we see is that removing orography actually reduces the variability in meridional mode too, which is anti-symmetric regarding equator. But it increases variability in odd modes, those that are symmetric regarding equator. This actually shows that the hemispheric differences are reduced. And the other thing is that with removing orography we have more variability in meridional mode. Three, zonal wave number four. This is actually just the pattern of a Rosby wave train in the mid-latitudes with the zonal wave number four. This is where the most variability is in when we remove orography. So I presented a novel method for the decomposition of 3D global circulation to the balanced and unbalanced component. And it has been applied to diagnose a global response to tropical heating perturbations. In tropics the balanced and unbalanced components of the response have comparable amplitudes. Circulation response is confined close to the heating source, which is possibly due to the dipole structure of the heating. And the mid-latitudinal response consists of balanced Rosby wave train. Removal of orography increases the variability in symmetric Rosby modes and the largest variability appears in Rosby modes three at zonal wave number four.