Nearly all natural phenomena are studied by collecting data, creating models and evaluating their efficiency. Thus, the outcomes of data analyses are strongly dependent on model selection, as well as the choice of the inference methods. However, as Karl Popper suggested, there are only two kinds of theories: those that are wrong and those that are incomplete. Theories and models usually have limited explanation power and natural processes are almost never adequately or completely modeled. The remaining unexplained fluctuations are usually considered as “noise”, in contrast to the deterministic “signal”. This signal-to-noise dichotomy becomes quite problematic when noise levels are close to those of deterministic variations.

In climate science, this has been demonstrated by Hasselmann [1], who explained how short-time-scale phenomena, modeled as stochastic perturbations, could affect long-term climate variations. Recently, his proposal for a stochastic approach to climate modelling came gain to the spotlight, as it could potential be combined and improve the physically-based numerical models [2]. Another scientific discipline, in which stochastic modelling has been extensively used is hydrology [3]. Among its numerous applications it has been used to investigate how complex phenomena behave across temporal scales [4]. The latter has been the central theme of my research, since the beginning of my PhD in stochastic hydroclimatology.

My main methodological toolbox contains:

  • Stochastic methods for uncertainty analysis and simulation [5, 6].
  • Data-mining approaches for dimensionality reduction and visualization [7].
  • Machine learning techniques for classification and forecasting [5].

The overarching theme in these analyses is the determination of multi-scale variability in the global hydrological cycle, which is my main short-term scientific goal. To achieve this, I combine the state-of-the-art observational products and palaeoclimatic reconstructions to create a robust description of the global water cycle across the spatio-temporal scale continuum [8, 9]. An important step in the development of an inclusive methodological framework, has been already made with open international collaborations [10, 11], which led to the development of an R package and a series of on-going workshops. Thus, at the moment I am working into the quantification on Earth’s water cycle fluctuations and their relationship to global temperature. My research plan is to develop stochastic 1- and/or 2-D models, derived from the physical laws in order to determine Earth’s water cycle sensitivity. The estimation of its sensitivity, as well as its rate of change, could in turn provide the range of potential atmospheric/climatic regimes.

Earth is not the only domain that observational evidence is rapidly increasing. The same holds true for exoplanetary research, due to the contribution of NASA’s Kepler mission and is expected to become even larger with the long-anticipated TESS and JWST missions. Both general circulation climate models and palaeoclimatic representations have already been suggested as potential approaches in exoplanetary research [12, 13]. I am really excited to see these approaches applied in the representation of climatic analogues for terrestrial exosolar planets. This will help to investigate whether exosolar planets are able to develop and maintain an active water cycle, in an attempt to constrain more the habitable zone to the subsection were terrestrial planets with a hydrological water cycle might be discovered.

Starting from the Earth and its water cycle, we can look for the structural components that other planets should have in order to preserve similar climatic conditions. On the other hand, the knowledge stemming from exoplanets and astronomical research in general, can help us to broaden our understanding of the potential outcomes of the recent abrupt rise in global temperature. In both cases the starting point and the destination are common: our unique planet.


[1] Hasselmann (1976) Tellus, 28(6), 473-485 [2] Lucarini et al. (2014) Reviews of Geophysics 52(4) 829-859 [3] Bras & Rodríguez-Iturbe (1985) Courier Corporation [4] Koutsoyiannis (2003) Hydrological Sciences Journal 48(1) 3-24 [5] Markonis, Y., et al. (2018) Advances in Water Resources, DOI: 10.1016/j.advwatres.2018.01.003 [6] Markonis, Y., et al. Scientific Reports, in revision [7] Markonis Y. et al. Nature Communications, 2018. [8] Markonis, Y., & Koutsoyiannis, D. (2013) Surveys in Geophysics, 34(2), 181-207. [9] Markonis, Y., & Koutsoyiannis, D. (2016) Nature Climate Change, 6(4), 399-401 [10] Pappas, C., et al.  Nature Ecology & Evolution, 1(9), 1263 [11] Papalexiou, S. M., et al., in preparation [12] Shields, A. L., et al. (2013) Astrobiology, 13(8), 715-739 [13] Goldblatt, C., Kavanagh, L., & Dewey, M. (2017) Geoscientific Model Development, 10(12), 3931