Deep Convolutional Neural Network (CNN) is a special type of Neural Networks, which has shown exemplary performance on several competitions related to Computer Vision and Image Processing. Some of the exciting application areas of CNN include Image Classification and Segmentation, Object Detection, Video Processing, Natural Language Processing, and Speech Recognition. The powerful learning ability of deep CNN is primarily due to the use of multiple feature extraction stages that can automatically learn representations from the data. The availability of a large amount of data and improvement in the hardware technology has accelerated the research in CNNs, and recently interesting deep CNN architectures have been reported. Several inspiring ideas to bring advancements in CNNs have been explored, such as the use of different activation and loss functions, parameter optimization, regularization, and architectural innovations. However, the significant improvement in the representational capacity of the deep CNN is achieved through architectural innovations. Notably, the ideas of exploiting spatial and channel information, depth and width of architecture, and multi-path information processing have gained substantial attention. Similarly, the idea of using a block of layers as a structural unit is also gaining popularity. This survey thus focuses on the intrinsic taxonomy present in the recently reported deep CNN architectures and, consequently, classifies the recent innovations in CNN architectures into seven different categories. These seven categories are based on spatial exploitation, depth, multi-path, width, feature-map exploitation, channel boosting, and attention. Additionally, the elementary understanding of CNN components, current challenges, and applications of CNN are also provided.

https://link.springer.com/content/pdf/10.1007/s10462-020-09825-6.pdf

A survey of the recent architectures of deep convolutional neural networks

List of illustrations Some useful constants and units Introduction 1. The facts which call for explanation. A sketch of the surface distribution of the meteorological elements over the globe 2. Some statistical and thermal relationships 3. Water-vapour in the atmosphere. The ascent of damp air 4. Thermodynamics of the atmosphere 5. Radiation 6. Radiation in the troposphere 7. Radiative equilibrium and the stratosphere 8. The general equations of motion 9. Motion under balanced forces: the gradient wind 10. Surfaces of discontinuity 11. The general aspects of turbulence 12. Turbulence in the atmosphere: the eddies as diffusing agencies 13. The classification of winds 14. The transformations of energy in the atmosphere 15. The growth of cyclic circulations and of pressure inequalities in the atmosphere 16. The idea of air masses 17. The Polar front and its relation to the development of cyclones 18. Anticyclones 19. The general circulation of the atmosphere Appendix Index of names Index of subjects.

Physical and Dynamical Meteorolgy. By D. Brunt. Second edition. Pp. xxiv, 428. 25s. 1939. (Cambridge)

In order to estimate potential greenhouse gas flux rates from soils under different land use and climate, and to particularly assess the influence of soil temperature and soil moisture, we measured fluxes of nitrous oxide (N2O), nitric oxide (NO), carbon dioxide (CO2) and methane (CH4) from intact soil cores obtained from 13 European sites under controlled laboratory conditions. The soils covered the different climates of Europe and included four different land-use types: croplands, forests, grasslands and wetlands. In a two-factorial experimental design, the soil cores were incubated under four temperatures (5–20°C) and water contents (20–80% water-filled pore space). We found a non-linear increase of N2O, NO and CO2 emissions with increasing temperature. Nitrous oxide emissions were positively correlated with soil moisture, while NO emission and CH4 oxidation rates were negatively correlated with soil moisture. Maximum CO2 emissions occurred at intermediate soil moisture.

Different land-use types strongly affected greenhouse gas fluxes. Nitrous oxide and CO2 emissions were highest in grassland soils, while NO emissions were highest in forest soils. In grasslands, high soil microbial activity stimulated by high carbon (C) and nitrogen (N) contents, dense root systems and high C input from above-ground decaying biomass was the most likely cause for high N2O and CO2 emissions. High NO emissions from forest soils were mainly attributed to low pH and high soil porosity. Northern soils showed the greatest capacity to take up CH4 under warmer and dryer soil conditions. Nitric oxide emissions were positively correlated with N input.

Greenhouse gas emissions from European soils under different land use: effects of soil moisture and temperature

Nitrifiers and denitrifiers respond rapidly to changed moisture and increasing temperature in a pristine forest soil.

Measurements of the production and consumption of organic material have been a focus of aquatic science for more than 80 years. Over the last century, a variety of approaches have been developed and employed for measuring rates of gross primary production (Pg), respiration (R), and net ecosystem production (Pn = Pg − R) within aquatic ecosystems. Here, we reconsider the range of approaches and applications for ecosystem metabolism measurements, and suggest ways by which such studies can continue to contribute to aquatic ecology. This paper reviews past and contemporary studies of aquatic ecosystem-level metabolism to identify their role in understanding and managing aquatic systems. We identify four broad research objectives that have motivated ecosystem metabolism studies: (1) quantifying magnitude and variability of metabolic rates for cross-system comparison, (2) estimating organic matter transfer between adjacent systems or subsystems, (3) measuring ecosystem-scale responses to perturbation, both natural and anthropogenic, and (4) quantifying and calibrating models of biogeochemical processes and trophic networks. The magnitudes of whole-system gross primary production, respiration and net ecosystem production rates vary among aquatic environments and are partly constrained by the chosen methodology. We argue that measurements of ecosystem metabolism should be a vital component of routine monitoring at larger scales in the aquatic environment using existing flexible, precise, and durable sensor technologies. Current and future aquatic ecosystem studies will benefit from application of new methods for metabolism measurements, which facilitate integration of process measurements and calibration of models for addressing fundamental questions involving ecosystem-scale processes.

/pdf/the-metabolism-of-aquatic-ecosystems-history-applications-34azmlnmcg.pdf

The metabolism of aquatic ecosystems: history, applications, and future challenges

To improve our understanding of tidal sandbank dynamics, we have developed a
nonlinear morphodynamic model. A crucial property of the model is that it fully
resolves the dynamics on the fast (tidal) timescale, allowing for asymmetric tidal flow with an M0, M2, and M4 component. This approach, extending earlier research on the formation of tidal sandbanks, leads to equilibrium profiles. Their heights (60-90% of the water depth) and shapes are controlled by the mode of sediment transport and the hydrodynamic conditions. Bed load transport under symmetrical tidal conditions leads to high spiky banks. Several mechanisms tend to lower and smooth these profiles, such as the relaxation of suspended sediment, wind wave stirring, and tidal asymmetry. This last causes the profiles to be asymmetric, as well. The morphodynamic equilibrium expresses a tidally averaged balance between a destabilizing flux due to fluid drag and the downslope transport induced by both tidal flow and wind wave stirring. The modeled profiles are in fair agreement with observations from the North Sea.

/pdf/the-cross-sectional-shape-of-tidal-sandbanks-modeling-and-2cxjhignam.pdf

The cross-sectional shape of tidal sandbanks: modeling and observations

A new method to calculate flow in compound channels is proposed: the interacting divided channel method (IDCM), based on a new parametrization of the interface stress between adjacent flow compartments, typically between the main channel and floodplain of a two-stage channel. This expression is motivated by scaling arguments and allows for a simple analytical solution of the average flow velocities in different compartments. Good agreement is found between the analytical model results and laboratory data from the literature.

Interacting divided channel method for compound channel flow

https://ris.utwente.nl/ws/files/6565138/morphodynamics_of_tidal.pdf

The morphodynamics of tidal sand waves: A model overview

River deltas and individual delta lobes frequently face reduction of sediment supply, either from the geologic process of river avulsion or, more recently, due to human activities such as river damming. Using a process-based shoreline evolution model, we investigate wave reworking of delta shorelines after fluvial input elimination. Model results suggest that littoral sediment transport can result in four characteristic modes of delta abandonment, ranging from diffusional smoothing of the delta (or delta lobe) to the development of recurved spits. A straightforward analysis of delta shape and wave characteristics provides a framework for predicting the mode of delta abandonment. The observed morphologies of historically abandoned delta lobes, including those of the Nile, Ebro, and Rhone rivers, fit within this framework. Our results provide quantitative insight into the potential evolution of active delta environments in light of future extreme reduction of fluvial sediment input.

https://agupubs.onlinelibrary.wiley.com/doi/pdf/10.1002/2013GL058231

Wave reworking of abandoned deltas

An idealized model for tide propagation and amplification in semi-enclosed rectangular basins is presented, accounting for depth differences by a combination of longitudinal and lateral topographic steps. The basin geometry is formed by several adjacent compartments of identical width, each having either a uniform depth or two depths separated by a transverse topographic step. The problem is forced by an incoming Kelvin wave at the open end, while allowing waves to radiate outward. The solution in each compartment is written as the superposition of (semi)-analytical wave solutions in an infinite channel, individually satisfying the depth-averaged linear shallow water equations on the f plane, including bottom friction. A collocation technique is employed to satisfy continuity of elevation and flux across the longitudinal topographic steps between the compartments. The model results show that the tidal wave in shallow parts displays slower propagation, enhanced dissipation and amplified amplitudes. This reveals a resonance mechanism, occurring when the length of the shallow end is roughly an odd multiple of the quarter Kelvin wavelength. Alternatively, for sufficiently wide basins, also Poincare waves may become resonant. A transverse step implies different wavelengths of the incoming and reflected Kelvin wave, leading to increased amplitudes in shallow regions and a shift of amphidromic points in the direction of the deeper part. Including the shallow parts near the basin’s closed end (thus capturing the Kelvin resonance mechanism) is essential to reproduce semi-diurnal and diurnal tide observations in the Gulf of California, the Adriatic Sea and the Persian Gulf.

/pdf/influence-of-topography-on-tide-propagation-and-3u80hvdtjn.pdf

Pieter C. Roos

Papers

The cross-sectional shape of tidal sandbanks: modeling and observations

Interacting divided channel method for compound channel flow

The morphodynamics of tidal sand waves: A model overview

Wave reworking of abandoned deltas

Influence of topography on tide propagation and amplification in semi-enclosed basins