@article{Ayzel2021,
  author    = {Ayzel, Georgy},
  title     = {Deep neural networks in hydrology},
  series = {Vestnik of Saint Petersburg University. Earth Sciences},
  volume    = {66},
  journal   = {Vestnik of Saint Petersburg University. Earth Sciences},
  number    = {1},
  publisher = {Univ. Press},
  address   = {St. Petersburg},
  issn      = {2541-9668},
  doi       = {10.21638/spbu07.2021.101},
  pages     = {5 -- 18},
  year      = {2021},
  abstract  = {For around a decade, deep learning - the sub-field of machine learning that refers to artificial neural networks comprised of many computational layers - modifies the landscape of statistical model development in many research areas, such as image classification, machine translation, and speech recognition. Geoscientific disciplines in general and the field of hydrology in particular, also do not stand aside from this movement. Recently, the proliferation of modern deep learning-based techniques and methods has been actively gaining popularity for solving a wide range of hydrological problems: modeling and forecasting of river runoff, hydrological model parameters regionalization, assessment of available water resources. identification of the main drivers of the recent change in water balance components. This growing popularity of deep neural networks is primarily due to their high universality and efficiency. The presented qualities, together with the rapidly growing amount of accumulated environmental information, as well as increasing availability of computing facilities and resources, allow us to speak about deep neural networks as a new generation of mathematical models designed to, if not to replace existing solutions, but significantly enrich the field of geophysical processes modeling. This paper provides a brief overview of the current state of the field of development and application of deep neural networks in hydrology. Also in the following study, the qualitative long-term forecast regarding the development of deep learning technology for managing the corresponding hydrological modeling challenges is provided based on the use of "Gartner Hype Curve", which in the general details describes a life cycle of modern technologies.},
  language  = {en}
}
@article{HeckenbachBruneGlerumetal.2021,
  author    = {Heckenbach, Esther Lina and Brune, Sascha and Glerum, Anne C. and Bott, Judith},
  title     = {Is there a speed limit for the thermal steady-state assumption in continental rifts?},
  series = {Geochemistry, geophysics, geosystems : G 3 ; an electronic journal of the earth sciences},
  volume    = {22},
  journal   = {Geochemistry, geophysics, geosystems : G 3 ; an electronic journal of the earth sciences},
  number    = {3},
  publisher = {Wiley},
  address   = {Hoboken, NJ},
  issn      = {1525-2027},
  doi       = {10.1029/2020GC009577},
  pages     = {18},
  year      = {2021},
  abstract  = {The lithosphere is often assumed to reside in a thermal steady-state when quantitatively describing the temperature distribution in continental interiors and sedimentary basins, but also at active plate boundaries. Here, we investigate the applicability limit of this assumption at slowly deforming continental rifts. To this aim, we assess the tectonic thermal imprint in numerical experiments that cover a range of realistic rift configurations. For each model scenario, the deviation from thermal equilibrium is evaluated. This is done by comparing the transient temperature field of every model to a corresponding steady-state model with an identical structural configuration. We find that the validity of the thermal steady-state assumption strongly depends on rift type, divergence velocity, sampling location, and depth within the rift. Maximum differences between transient and steady-state models occur in narrow rifts, at the rift sides, and if the extension rate exceeds 0.5-2 mm/a. Wide rifts, however, reside close to thermal steady-state even for high extension velocities. The transient imprint of rifting appears to be overall negligible for shallow isotherms with a temperature less than 100 degrees C. Contrarily, a steady-state treatment of deep crustal isotherms leads to an underestimation of crustal temperatures, especially for narrow rift settings. Thus, not only relatively fast rifts like the Gulf of Corinth, Red Sea, and Main Ethiopian Rift, but even slow rifts like the Kenya Rift, Rhine Graben, and Rio Grande Rift must be expected to feature a pronounced transient component in the temperature field and to therefore violate the thermal steady-state assumption for deeper crustal isotherms.},
  language  = {en}
}