Endre søk
Begrens søket
1 - 10 of 10
RefereraExporteraLink til resultatlisten
Permanent link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf
Treff pr side
  • 5
  • 10
  • 20
  • 50
  • 100
  • 250
Sortering
  • Standard (Relevans)
  • Forfatter A-Ø
  • Forfatter Ø-A
  • Tittel A-Ø
  • Tittel Ø-A
  • Type publikasjon A-Ø
  • Type publikasjon Ø-A
  • Eldste først
  • Nyeste først
  • Skapad (Eldste først)
  • Skapad (Nyeste først)
  • Senast uppdaterad (Eldste først)
  • Senast uppdaterad (Nyeste først)
  • Disputationsdatum (tidligste først)
  • Disputationsdatum (siste først)
  • Standard (Relevans)
  • Forfatter A-Ø
  • Forfatter Ø-A
  • Tittel A-Ø
  • Tittel Ø-A
  • Type publikasjon A-Ø
  • Type publikasjon Ø-A
  • Eldste først
  • Nyeste først
  • Skapad (Eldste først)
  • Skapad (Nyeste først)
  • Senast uppdaterad (Eldste først)
  • Senast uppdaterad (Nyeste først)
  • Disputationsdatum (tidligste først)
  • Disputationsdatum (siste først)
Merk
Maxantalet träffar du kan exportera från sökgränssnittet är 250. Vid större uttag använd dig av utsökningar.
  • 1. de Vries, Hylke
    et al.
    Scher, Sebastian
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Haarsma, Rein
    Drijfhout, Sybren
    van Delden, Aarnout
    How Gulf-Stream SST-fronts influence Atlantic winter storms: Results from a downscaling experiment with HARMONIE to the role of modified latent heat fluxes and low-level baroclinicity2019Inngår i: Climate Dynamics, ISSN 0930-7575, E-ISSN 1432-0894, Vol. 52, nr 9-10, s. 5899-5909Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    The strong horizontal gradients in sea surface temperature (SST) of the Atlantic Gulf Stream exert a detectable influence on extratropical cyclones propagating across the region. This is shown in a sensitivity experiment where 24 winter storms taken from ERA-Interim are simulated with HARMONIE at 10-km resolution. Each storm is simulated twice. First, using observed SST (REF). In the second simulation a smoothed SST is offered (SMTH), while lateral and upper-level boundary conditions are unmodified. Each storm pair propagates approximately along the same track, however their intensities (as measured by maximal near-surface wind speed or 850-hPa relative vorticity) differ up to +/- 25%. A 30-member ensemble created for one of the storms shows that on a single-storm level the response is systematic rather than random. To explain the broad response in storm strength, we show that the SST-adjustment modifies two environmental parameters: surface latent heat flux (LHF) and low-level baroclinicity (B). LHF influences storms by modifying diabatic heating and boundary-layer processes such as vertical mixing. The position of each storm's track relative to the SST-front is important. South of the SST-front the smoothing leads to lower SST, reduced LHF and storms with generally weaker maximum near-surface winds. North of the SST-front the increased LHF tend to enhance the winds, but the accompanying changes in baroclinicity are not necessarily favourable. Together these mechanisms explain up to 80% of the variability in the near-surface maximal wind speed change. Because the mechanisms are less effective in explaining more dynamics-oriented indicators like 850 hPa relative vorticity, we hypothesise that part of the wind-speed change is related to adjustment of the boundary-layer processes in response to the LHF and B changes.

  • 2.
    Scher, Sebastian
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Artificial intelligence in weather and climate prediction: Learning atmospheric dynamics2020Doktoravhandling, med artikler (Annet vitenskapelig)
    Abstract [en]

    Weather and climate prediction is dominated by high dimensionality, interactions on many different spatial and temporal scales and chaotic dynamics. This makes many problems in the field quite complex ones, and also state-of-the-art numerical models are - despite their immense computational costs - not sufficient for many applications. Therefore, it is appealing to use emerging new technologies such as artificial intelligence to tackle these problems.

    We show that it is possible to use deep neural networks to emulate the full dynamics of a strongly simplified general circulation model, providing both good forecasts of the model state several days ahead as well as stable long-term climate timeseries. This method partly also works on more complex and realistic models, but only for forecasting the model's weather several days ahead, not for creating climate runs. It is sufficient to use 50-100 years of data for training the networks. The same neural network method can be combined with singular value decomposition from numerical ensemble weather forecasting in order to generate probabilistic ensemble forecasts with the neural networks.

    On a more fundamental level, we show that in a simple dynamical systems setting there seem to be limitations in the ability of feed-forward neural networks to generalize to new regions of the system. This is caused by different parts of the network learning to model different parts of the system. Contradictory, for another simple dynamical system this is shown not to be an issue, raising doubts on the usefulness of results from simple models in the context of more complex ones. Additionally, we show that neural networks are to some extent able to “learn” the influence of slowly changing external forcings on the dynamics of the system, but only given broad enough forcing regimes.

    Finally, we present a method to complement operational weather forecasts. Given the initial fields and the error of past weather forecasts, a neural network is used to predict the uncertainty in new forecasts, given only the initial field of the new forecast.

    Fulltekst (pdf)
    Artificial intelligence in weather and climate prediction: Learning atmospheric dynamics
    Download (jpg)
    Omslagsframsida
  • 3.
    Scher, Sebastian
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Toward Data-Driven Weather and Climate Forecasting: Approximating a Simple General Circulation Model With Deep Learning2018Inngår i: Geophysical Research Letters, ISSN 0094-8276, E-ISSN 1944-8007, Vol. 45, nr 22, s. 12616-12622Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    It is shown that it is possible to emulate the dynamics of a simple general circulation model with a deep neural network. After being trained on the model, the network can predict the complete model state several time steps aheadwhich conceptually is making weather forecasts in the model world. Additionally, after being initialized with an arbitrary model state, the network can through repeatedly feeding back its predictions into its inputs create a climate run, which has similar climate statistics to the climate of the general circulation model. This network climate run shows no long-term drift, even though no conservation properties were explicitly designed into the network. Plain Language Summary Numerical weather prediction and climate models are complex computer programs that represent the physics of the atmosphere. They are essential tools for predicting the weather and for studying the Earth's climate. Recently, a lot of progress has been made in machine learning methods. These are data-driven algorithms that learn from existing data. We show that it is possible that such an algorithm learns the dynamics of a simple climate model. After being presented with enough data from the climate model, the network can successfully predict the time evolution of the model's state, thus replacing the dynamics of the model. This finding is an important step toward purely data-driven weather forecastingthus weather forecasting without the use of traditional numerical models and also opens up new possibilities for climate modeling.

  • 4.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Messori, Gabriele
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU). Uppsala University, Sweden.
    Ensemble neural network forecasts with singular value decompositionManuskript (preprint) (Annet vitenskapelig)
    Abstract [en]

    Ensemble weather forecasts enable a measure of uncertainty to be attached to each forecast by computing the ensemble's spread. However, generating an ensemble with a good error-spread relationship is far from trivial, and a wide range of approaches to achieve this have been explored. Random perturbations of the initial model state typically provide unsatisfactory results when applied to numerical weather prediction models. Singular value decomposition has proved more successful in this context, and as a result has been widely used for creating perturbed initial states of weather prediction models. We demonstrate how to apply the technique of singular value decomposition to purely neural-network based forecasts. Additionally, we explore the use of random initial perturbations for neural network ensembles, and the creation of neural network ensembles via retraining the network. We find that the singular value decomposition results in ensemble forecasts that have some probabilistic skill, but are inferior to the ensemble created by retraining the neural network several times. Compared to random initial perturbations, the singular value technique performs better when forecasting a simple general circulation model, comparably when forecasting atmospheric reanalysis data, and worse when forecasting the lorenz95 system - a highly idealized model designed to mimic certain aspects of the mid-latitude atmosphere.

  • 5.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU). Uppsala University, Sweden.
    Messori, Gabriele
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Generalization properties of feed-forward neural networks trained on Lorenz systems2019Inngår i: Nonlinear processes in geophysics, ISSN 1023-5809, E-ISSN 1607-7946, Vol. 26, nr 4, s. 381-399Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Neural networks are able to approximate chaotic dynamical systems when provided with training data that cover all relevant regions of the system's phase space. However, many practical applications diverge from this idealized scenario. Here, we investigate the ability of feed-forward neural networks to (1) learn the behavior of dynamical systems from incomplete training data and (2) learn the influence of an external forcing on the dynamics. Climate science is a real-world example where these questions may be relevant: it is concerned with a non-stationary chaotic system subject to external forcing and whose behavior is known only through comparatively short data series. Our analysis is performed on the Lorenz63 and Lorenz95 models. We show that for the Lorenz63 system, neural networks trained on data covering only part of the system's phase space struggle to make skillful short-term forecasts in the regions excluded from the training. Additionally, when making long series of consecutive forecasts, the networks struggle to reproduce trajectories exploring regions beyond those seen in the training data, except for cases where only small parts are left out during training. We find this is due to the neural network learning a localized mapping for each region of phase space in the training data rather than a global mapping. This manifests itself in that parts of the networks learn only particular parts of the phase space. In contrast, for the Lorenz95 system the networks succeed in generalizing to new parts of the phase space not seen in the training data. We also find that the networks are able to learn the influence of an external forcing, but only when given relatively large ranges of the forcing in the training. These results point to potential limitations of feed-forward neural networks in generalizing a system's behavior given limited initial information. Much attention must therefore be given to designing appropriate train-test splits for real-world applications.

  • 6.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Messori, Gabriele
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU). Uppsala University, Sweden.
    How Global Warming Changes the Difficulty of Synoptic Weather Forecasting2019Inngår i: Geophysical Research Letters, ISSN 0094-8276, E-ISSN 1944-8007, Vol. 46, nr 5, s. 2931-2939Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Global warming projections point to a wide range of impacts on the climate system, including changes in storm track activity and more frequent and intense extreme weather events. Little is however known on whether and how global warming may affect the atmosphere's predictability and thus our ability to produce accurate weather forecasts. Here, we combine a state-of-the-art climate and a state-of-the-art ensemble weather prediction model to show that, in a business-as-usual 21st century setting, global warming could significantly change the predictability of the atmosphere, defined here via the expected error of weather predictions. Predictability of synoptic weather situations could significantly increase, especially in the Northern Hemisphere. This can be explained by a decrease in the meridional temperature gradient. Contrarily, summertime predictability of weekly rainfall sums might significantly decrease in most regions. Plain Language Summary Due to the chaotic nature of the atmosphere, it is impossible to make weather forecasts that are completely accurate. Therefore, all weather forecasts are inherently uncertain to a certain degree. However, this uncertainty-and thus the difficulty of making good forecastsis not the same for all forecasts. This opens up the highly important question whether global warming will affect the difficulty of weather forecasts. Due to the enormous socioeconomic importance of accurate weather forecasts, it is essential to know whether climate change adaption policies also need to take into account potential changes in the difficulty and accuracy of weather forecasts. We show that in a warmer world, it will be easier to predict fields such as temperature and pressure. Contrarily, it will be harder to make accurate precipitation forecasts, which might strongly affect both disaster prevention and rainfall-dependent industries such as the energy sector, all of which heavily rely on accurate precipitation forecasts. Additionally, we show that the uncertainty of predictions of pressure fields is to a large extent controlled by fluctuations in the temperature difference between the North Pole and the equator. This is a new and important insight into the fundamentals of weather forecast uncertainty.

  • 7.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Messori, Gabriele
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Predicting weather forecast uncertainty with machine learning2018Inngår i: Quarterly Journal of the Royal Meteorological Society, ISSN 0035-9009, E-ISSN 1477-870X, Vol. 144, nr 717, s. 2830-2841Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Weather forecasts are inherently uncertain. Therefore, for many applications forecasts are only considered valuable if an uncertainty estimate can be assigned to them. Currently, the best method to provide a confidence estimate for individual forecasts is to produce an ensemble of numerical weather simulations, which is computationally very expensive. Here, we assess whether machine learning techniques can provide an alternative approach to predict the uncertainty of a weather forecast given the large-scale atmospheric state at initialization. We propose a method based on deep learning with artificial convolutional neural networks that is trained on past weather forecasts. Given a new weather situation, it assigns a scalar value of confidence to medium-range forecasts initialized from the said atmospheric state, indicating whether the predictability is higher or lower than usual for the time of the year. While our method has a lower skill than ensemble weather forecast models in predicting forecast uncertainty, it is computationally very efficient and outperforms a range of alternative methods that do not involve performing numerical forecasts. This shows that it is possible to use machine learning in order to estimate future forecast uncertainty from past forecasts. The main constraint in the performance of our method seems to be the number of past forecasts available for training the machine learning algorithm.

  • 8.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Messori, Gabriele
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Selective ensemble mean technique for severe European windstorms2019Inngår i: Quarterly Journal of the Royal Meteorological Society, ISSN 0035-9009, E-ISSN 1477-870X, Vol. 145, nr 718, s. 376-385Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    We show that ensemble forecasts of extreme European windstorms can be improved up to a lead time of 36-48 hr by sub-selecting ensemble members based on their performance at very short lead times (up to 12 hr). This applies to both the ensemble mean position of the cyclone centre and the ensemble windstorm footprint over the continent. A number of ensemble forecasts, including those from the ECMWF Ensemble Prediction System, are initialised every 12 hr and disseminated several hours after initialisation; therefore our approach has the potential to provide improved forecasts in an operational context. The analysis is performed on GEFS reforecast data.

  • 9.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Messori, Gabriele
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU). Uppsala University, Sweden.
    Weather and climate forecasting with neural networks: using general circulation models (GCMs) with different complexity as a study ground2019Inngår i: Geoscientific Model Development, ISSN 1991-959X, E-ISSN 1991-9603, Vol. 12, nr 7, s. 2797-2809Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Recently, there has been growing interest in the possibility of using neural networks for both weather forecasting and the generation of climate datasets. We use a bottom-up approach for assessing whether it should, in principle, be possible to do this. We use the relatively simple general circulation models (GCMs) PUMA and PLASIM as a simplified reality on which we train deep neural networks, which we then use for predicting the model weather at lead times of a few days. We specifically assess how the complexity of the climate model affects the neural network's forecast skill and how dependent the skill is on the length of the provided training period. Additionally, we show that using the neural networks to reproduce the climate of general circulation models including a seasonal cycle remains challenging - in contrast to earlier promising results on a model without seasonal cycle.

  • 10.
    Scher, Sebastian
    et al.
    Stockholms universitet, Naturvetenskapliga fakulteten, Meteorologiska institutionen (MISU).
    Molinder, Jennie
    Machine Learning-Based Prediction of Icing-Related Wind Power Production Loss2019Inngår i: IEEE Access, E-ISSN 2169-3536, Vol. 7, s. 129421-129429Artikkel i tidsskrift (Fagfellevurdert)
    Abstract [en]

    Ice-growth on wind-turbines can lead to a large reduction of energy production. Since ice-growth on the turbines is not part of standard weather prediction data, forecasts of power production can have large errors when ice-growth occurs. We propose a statistical method based on random-forest regression to predict the production loss induced by ice-growth. It takes as input both regional weather forecasts and on-site measurements, and predicts relative power production loss up to 42 hours ahead in order to improve the prediction for the next-day energy production. The method is trained on past forecasts and measurements, and significantly outperforms a simple - but also useful - persistence baseline especially at longer lead times. It reduces the absolute error of production forecasts by similar to 100kW and is comparable in skill to physics-based icing models. The weather prediction data is the most important input for the statistical predictions, and on-site measurements are not absolutely necessary. The algorithm is computationally very inexpensive and can easily be retrained for every new forecast.

1 - 10 of 10
RefereraExporteraLink til resultatlisten
Permanent link
Referera
Referensformat
  • apa
  • ieee
  • modern-language-association-8th-edition
  • vancouver
  • Annet format
Fler format
Språk
  • de-DE
  • en-GB
  • en-US
  • fi-FI
  • nn-NO
  • nn-NB
  • sv-SE
  • Annet språk
Fler språk
Utmatningsformat
  • html
  • text
  • asciidoc
  • rtf