Signatures of geostrophic turbulence in power spectra and third-order structure function of offshore wind speed fluctuations.

We analyze offshore wind speeds with a time resolution of one second over a long period of 20 months for different heights above the sea level. Energy spectra extending over more than seven decades give a comprehensive picture of wind fluctuations, including intermittency effects at small length scales and synoptic weather phenomena at large scales. The spectra S(f) show a scaling behavior consistent with three-dimensional turbulence at high frequencies f, followed by a regime at lower frequencies, where fS(f) varies weakly. Lowering the frequency below a crossover frequency [Formula: see text], a rapid rise of fS(f) occurs. An analysis of the third-order structure function [Formula: see text] of wind speed differences for a given time lag [Formula: see text] shows a rapid change from negative to positive values of [Formula: see text] at [Formula: see text]. Remarkably, after applying Taylor's hypothesis locally, we find the third-order structure function to exhibit a behavior very similar to that obtained previously from aircraft measurements at much higher altitudes in the atmosphere. In particular, the third-order structure function grows linearly with the separation distance for negative [Formula: see text], and with the third power for positive [Formula: see text]. This allows us to estimate energy and enstrophy dissipation rates for offshore wind. The crossover from negative to positive values occurs at about the same separation distance of 400 km as found from the aircraft measurements, suggesting that this length is independent of the altitude in the atmosphere.

Generation of Atmospheric Turbulence with Unprecedentedly Large Reynolds Number in a Wind Tunnel.

Generating laboratory flows resembling atmospheric turbulence is of prime importance to study the effect of wind fluctuations on objects such as buildings, vehicles, or wind turbines. A novel driving of an active grid following a stochastic process is used to generate velocity fluctuations with correlation lengths, and, thus, integral scales, much larger than the transverse dimension of the wind tunnel. The combined action of the active grid and a modulation of the fan speed allows one to generate a flow characterized by a four-decade inertial range and an integral scale Reynolds number of 2×10^{7}.

A topology-dynamics-based control strategy for multi-dimensional complex networked dynamical systems.

Complex systems are omnipresent and play a vital role in in our every-day lives. Adverse behavior of such systems has generated considerable interest in being able to control complex systems modeled as networks. Here, we propose a topology-dynamics-based approach for controlling complex systems modeled as networks of coupled multi-dimensional dynamical entities. For given dynamics and topology, we introduce an efficient scheme to identify in polynomial time a finite set of driver nodes, which - when endowed with the control function - steer the network to the desired behavior. We demonstrate the high suitability of our approach by controlling various networked multi-dimensional dynamics, coupled onto different topologies.

Heterogeneities in electricity grids strongly enhance non-Gaussian features of frequency fluctuations under stochastic power input.

Stochastic feed-in of fluctuating renewable energies is steadily increasing in modern electricity grids, and this becomes an important risk factor for maintaining power grid stability. Here, we study the impact of wind power feed-in on the short-term frequency fluctuations in power grids based on an Institute of Electrical and Electronics Engineers test grid structure, the swing equation for the dynamics of voltage phase angles, and a series of measured wind speed data. External control measures are accounted for by adjusting the grid state to the average power feed-in on a time scale of 1 min. The wind power is injected at a single node by replacing one of the conventional generator nodes in the test grid by a wind farm. We determine histograms of local frequencies for a large number of 1-min wind speed sequences taken from the measured data and for different injection nodes. These histograms exhibit a common type of shape, which can be described by a Gaussian distribution for small frequencies and a nearly exponentially decaying tail part. Non-Gaussian features become particularly pronounced for wind power injection at locations, which are weakly connected to the main grid structure. This effect is only present when taking into account the heterogeneities in transmission line and node properties of the grid, while it disappears upon homogenizing of these features. The standard deviation of the frequency fluctuations increases linearly with the average injected wind power.

Grand challenges in the science of wind energy.

Harvested by advanced technical systems honed over decades of research and development, wind energy has become a mainstream energy resource. However, continued innovation is needed to realize the potential of wind to serve the global demand for clean energy. Here, we outline three interdependent, cross-disciplinary grand challenges underpinning this research endeavor. The first is the need for a deeper understanding of the physics of atmospheric flow in the critical zone of plant operation. The second involves science and engineering of the largest dynamic, rotating machines in the world. The third encompasses optimization and control of fleets of wind plants working synergistically within the electricity grid. Addressing these challenges could enable wind power to provide as much as half of our global electricity needs and perhaps beyond.

Propagation of wind-power-induced fluctuations in power grids.

Renewable generators perturb the electric power grid with heavily non-Gaussian and time correlated fluctuations. While changes in generated power on timescales of minutes and hours are compensated by frequency control measures, we report subsecond distribution grid frequency measurements with local non-Gaussian fluctuations which depend on the magnitude of wind power generation in the grid. Motivated by such experimental findings, we simulate the subsecond grid frequency dynamics by perturbing the power grid, as modeled by a network of phase coupled nonlinear oscillators, with synthetically generated wind power feed-in time series. We derive a linear response theory and obtain analytical results for the variance of frequency increment distributions. We find that the variance of short-term fluctuations decays, for large inertia, exponentially with distance to the feed-in node, in agreement with numerical results both for a linear chain of nodes and the German transmission grid topology. In sharp contrast, the kurtosis of frequency increments is numerically found to decay only slowly, not exponentially, in both systems, indicating that the non-Gaussian shape of frequency fluctuations persists over long ranges.

Detecting Hidden Units and Network Size from Perceptible Dynamics.

The number of units of a network dynamical system, its size, arguably constitutes its most fundamental property. Many units of a network, however, are typically experimentally inaccessible such that the network size is often unknown. Here we introduce a detection matrix that suitably arranges multiple transient time series from the subset of accessible units to detect network size via matching rank constraints. The proposed method is model-free, applicable across system types and interaction topologies, and applies to nonstationary dynamics near fixed points, as well as periodic and chaotic collective motion. Even if only a small minority of units is perceptible and for systems simultaneously exhibiting nonlinearities, heterogeneities, and noise, exact size detection is feasible. We illustrate applicability for a paradigmatic class of biochemical reaction networks.

Effects of particle locations on reconstructed particle images in digital holography.

The intensity and phase reconstructed from digital in-line holograms by the convolution approach are analyzed. Distortions of particle images depending on their position in the plane transverse to the optical axis are identified. For this purpose, the object fields of numerically simulated particle holograms as well as of experimental data are reconstructed. The results of three-dimensional correlations of numerical and experimental data are superior when the numerically generated reference volumes are adapted to the transverse locations of the particle. Thus, proof is given that the characteristics of a particle image change distinctly with the transverse position of the particle and that the numerical model successfully simulates these changes. Hence, this knowledge can be integrated in future particle position detection algorithms.

Particle depth position detection by 2D correlation in digital in-line holography.

In digital holographic particle image velocimetry, hologram truncation is a very prominent problem when the projection of the particle position to the sensor is close to the sensor edge. Using the convolution approach to reconstruct such a hologram yields a deformed particle image compared to a particle image resulting from a particle with a projection to the center of the sensor. This Letter shows that the deformation complicates particle position detection based on an algorithm originally developed for analog holography by Choo and Kang, and later applied to digital holography by Yang and Kang. This algorithm is refined for the detection of particle positions from the deformed images and applied to numerical and experimental data.

Disentangling the stochastic behavior of complex time series.

Complex systems involving a large number of degrees of freedom, generally exhibit non-stationary dynamics, which can result in either continuous or discontinuous sample paths of the corresponding time series. The latter sample paths may be caused by discontinuous events - or jumps - with some distributed amplitudes, and disentangling effects caused by such jumps from effects caused by normal diffusion processes is a main problem for a detailed understanding of stochastic dynamics of complex systems. Here we introduce a non-parametric method to address this general problem. By means of a stochastic dynamical jump-diffusion modelling, we separate deterministic drift terms from different stochastic behaviors, namely diffusive and jumpy ones, and show that all of the unknown functions and coefficients of this modelling can be derived directly from measured time series. We demonstrate appli- cability of our method to empirical observations by a data-driven inference of the deterministic drift term and of the diffusive and jumpy behavior in brain dynamics from ten epilepsy patients. Particularly these different stochastic behaviors provide extra information that can be regarded valuable for diagnostic purposes.

Granger-causality maps of diffusion processes.

Granger causality is a statistical concept devised to reconstruct and quantify predictive information flow between stochastic processes. Although the general concept can be formulated model-free it is often considered in the framework of linear stochastic processes. Here we show how local linear model descriptions can be employed to extend Granger causality into the realm of nonlinear systems. This novel treatment results in maps that resolve Granger causality in regions of state space. Through examples we provide a proof of concept and illustrate the utility of these maps. Moreover, by integration we convert the local Granger causality into a global measure that yields a consistent picture for a global Ornstein-Uhlenbeck process. Finally, we recover invariance transformations known from the theory of autoregressive processes.

Stochastic nature of series of waiting times.

Although fluctuations in the waiting time series have been studied for a long time, some important issues such as its long-range memory and its stochastic features in the presence of nonstationarity have so far remained unstudied. Here we find that the "waiting times" series for a given increment level have long-range correlations with Hurst exponents belonging to the interval 1/2

Turbulent character of wind energy.

Wind turbines generate electricity from turbulent wind. Large fluctuations, and, more importantly, frequent wind gusts cause a highly fluctuating electrical power feed into the grid. Such effects are the hallmark of high-frequency turbulence. Here we show evidence that it is the complex structure of turbulence that dominates the power output for one single wind turbine as well as for an entire wind farm. We illustrate the highly intermittent, peaked nature of wind power fed into the grid. Multifractal scaling is observed, as described initially by Kolmogorov's 1962 theory of turbulence. In parallel, we propose a stochastic model that converts wind speed signals into power output signals with appropriate multifractal statistics. As more and more wind turbines become integrated into our electric grids, a proper understanding of this intermittent power source must be worked out to ensure grid stability in future networks. Thus, our results stress the need for a profound understanding of the physics of turbulence and its impact on wind energy.

Markov properties in presence of measurement noise.

Recently, several powerful tools for the reconstruction of stochastic differential equations from measured data sets have been proposed [e.g., Siegert, Phys. Lett. A 243, 275 (1998); Hurn, J. Time Series Anal. 24, 45 (2003)]. Efficient application of the methods, however, generally requires Markov properties to be fulfilled. This constraint typically seems to be violated on small scales, which frequently is attributed to physical effects. On the other hand, measurement noise such as uncorrelated measurement and discretization errors has large impacts on the statistics of measurements on small scales. We demonstrate that the presence of measurement noise, likewise, spoils Markov properties of an underlying Markov process. This fact is promising for the further development of techniques for the reconstruction of stochastic processes from measured data, since limitations at small scales might stem from artificial noise sources rather than from intrinsic properties of the dynamics of the underlying process. Measurement noise, however, can be controlled much better than the intrinsic dynamics of the underlying process.

Reconstruction of complex dynamical systems affected by strong measurement noise.

This Letter reports on a new approach to properly analyze time series of dynamical systems which are spoilt by the simultaneous presence of dynamical noise and measurement noise. It is shown that even strong external measurement noise as well as dynamical noise which is an intrinsic part of the dynamical process can be quantified correctly, solely on the basis of measured time series and proper data analysis. Finally, real world data sets are presented pointing out the relevance of the new approach.

Increment definitions for scale-dependent analysis of stochastic data.

It is common for scale-dependent analysis of stochastic data to use the increment Delta(t,r) =xi(t+r)-xi(t) of a data set xi(t) as a stochastic measure, where r denotes the scale. For joint statistics of Delta(t,r) and Delta(t, r') the question of how to nest the increments on different scales r, r' is investigated. Here we show that in some cases spurious correlations between scales can be introduced by the common left-justified definition. The consequences for a Markov process are discussed. These spurious correlations can be avoided by an appropriate nesting of increments. We demonstrate this effect for different data sets and show how it can be detected and quantified. The problem allows to propose a unique method to distinguish between experimental data generated by a noiselike or a Langevin-like random-walk process, respectively.