GNSS reflectometry global ocean wind speed using deep learning: Development and assessment of CyGNSSnet

2022 ◽  
Vol 269 ◽  
pp. 112801
Author(s):  
Milad Asgarimehr ◽  
Caroline Arnold ◽  
Tobias Weigel ◽  
Chris Ruf ◽  
Jens Wickert
Keyword(s):  
Deep Learning ◽  
Wind Speed ◽  
Global Ocean ◽  
Gnss Reflectometry ◽  
Learning Development ◽  
Ocean Wind

Calibration and Cross Validation of Global Ocean Wind Speed Based on Scatterometer Observations

2020 ◽  
Vol 37 (2) ◽  
pp. 279-297 ◽  
Author(s):  
Agustinus Ribal ◽  
Ian R. Young
Keyword(s):  
Wind Speed ◽  
Cross Validation ◽  
Regression Line ◽  
Total Duration ◽  
Global Ocean ◽  
High Wind ◽  
Wind Speeds ◽  
Buoy Data ◽  
Ocean Wind

AbstractGlobal ocean wind speed observed from seven different scatterometers, namely, ERS-1, ERS-2, QuikSCAT, MetOp-A, OceanSat-2, MetOp-B, and Rapid Scatterometer (RapidScat) were calibrated against National Data Buoy Center (NDBC) data to form a consistent long-term database of wind speed and direction. Each scatterometer was calibrated independently against NDBC buoy data and then cross validation between scatterometers was performed. The total duration of all scatterometer data is approximately 27 years, from 1992 until 2018. For calibration purposes, only buoys that are greater than 50 km offshore were used. Moreover, only scatterometer data within 50 km of the buoy and for which the overpass occurred within 30 min of the buoy recording data were considered as a “matchup.” To carry out the calibration, reduced major axis (RMA) regression has been applied where the regression minimizes the size of the triangle formed by the vertical and horizontal offsets of the data point from the regression line and the line itself. Differences between scatterometer and buoy data as a function of time were investigated for long-term stability. In addition, cross validation between scatterometers and independent altimeters was also performed for consistency. The performance of the scatterometers at high wind speeds was examined against buoy and platform measurements using quantile–quantile (Q–Q) plots. Where necessary, corrections were applied to ensure scatterometer data agreed with the in situ wind speed for high wind speeds. The resulting combined dataset is believed to be unique, representing the first long-duration multimission scatterometer dataset consistently calibrated, validated and quality controlled.

NOC GNSS-R Global Ocean Wind Speed and Sea-Ice Products

2020 ◽  
Author(s):  
Matthew Hammond ◽  
Giuseppe Foti ◽  
Christine Gommenginger ◽  
Meric Srokosz ◽  
Martin Unwin ◽  
...  
Keyword(s):  
Wind Speed ◽  
Sea Ice ◽  
Global Ocean ◽  
Retrieval Algorithm ◽  
Navigation Satellite ◽  
Validation Data ◽  
Global Navigation ◽  
Ice Detection ◽  
Global Navigation Satellite ◽  
Ocean Wind

<p>Global Navigation Satellite System-Reflectometry (GNSS-R) is an innovative and rapidly developing approach to Earth Observation that makes use of signals of opportunity from Global Navigation Satellite Systems, which have been reflected off the Earth’s surface. Using GNSS-R data collected by the UK TechDemoSat-1 (TDS-1) between 2014 and 2018, the National Oceanography Centre (NOC) has developed a GNSS-R wind speed retrieval algorithm called the Calibrated Bistatic Radar Equation (C-BRE), which now features updated data quality control mechanisms including flagging of radio frequency interference (RFI) and sea-ice detection based on the GNSS-R waveform. Here we present an assessment of the performance of the latest NOC GNSS-R ocean wind speed and sea-ice products (NOC C-BRE v1.0) using validation data from the ECMWF ERA-5 re-analysis model output. Results show the capability of spaceborne GNSS-R sensors for accurate wind speed retrieval and sea-ice detection. Additionally, ground-processed Galileo returns collected by TDS-1 are examined and the geophysical sensitivity of reflected Galileo data to surface parameters is investigated. Preliminary results demonstrate the feasibility of spaceborne GNSS-R instruments receiving a combination of GNSS signals transmitted by multiple navigation systems, which offers the opportunity for frequent, high-quality ocean wind and sea-ice retrievals at low relative cost. Other advancements in GNSS-R technology are represented by future mission concepts such as HydroGNSS, a proposed ESA Scout mission opportunity by SSTL offering support for enhanced retrieval capabilities exploiting dual polarisation, dual frequency, and coherent reflected signal reception.</p>

Deep Learning for the derivation of GNSS Reflectometry global ocean wind speed

2021 ◽  
Author(s):  
Milad Asgarimehr ◽  
Caroline Arnold ◽  
Felix Stiehler ◽  
Tobias Weigel ◽  
Chris Ruf ◽  
...  
Keyword(s):  
Machine Learning ◽  
Remote Sensing ◽  
Deep Learning ◽  
Wind Speed ◽  
Sampling Rate ◽  
Machine Learning Techniques ◽  
Global Ocean ◽  
Deep Model ◽  
Learning Techniques ◽  
Unseen Data

<p>The Global Navigation Satellite System Reflectometry (GNSS-R) is a novel remote sensing technique exploiting GNSS signals after reflection off the Earth's surface. The capability of spaceborne GNSS-R to monitor ocean state and the surface wind is recently well demonstrated, which offers an unprecedented sampling rate and much robustness during rainfall. The Cyclone GNSS (CyGNSS) is the first spaceborne mission fully dedicated to GNSS-R, launched in December 2016.</p><p>Thanks to the low development costs of the GNSS-R satellite missions as well as the capability of tracking multiple reflected signals from numerous GNSS transmitters, the GNSS-R datasets are much bigger compared to those from conventional remote sensing techniques. The CyGNSS provides a high number of unique samples in the order of a few millions monthly.  Deep learning can therefore be implemented in GNSS-R even more efficiently than other remote sensing domains. With the upcoming GNSS-R CubeSats, the data volume is expected to increase in the near future and GNSS-R “Big data” can be a future challenge. Deep learning methods are additionally able to correct the potential effects, both technical and geophysical, dictated by data empirically when the mechanisms are not well described by the theoretical knowledge. This poses the question if GNSS-R should embrace deep learning and can benefit from this modern data scientific method like other Earth Observation domains.</p><p>The receivers onboard CyGNSS cross-correlate the reflected signals received at a nadir antenna to a locally generated replica. The cross-correlation power at a range of the signal delay and Doppler frequency shift is the observational output of the receivers being called delay-Doppler Maps (DDMs). The mapped power is inversely proportional to the ocean roughness and consequently surface winds.</p><p>Few recent studies innovatively show some merits of machine learning techniques for the derivations of ocean winds from the DDMs. However, the capability of machine learning techniques, especially deep learning for an operational data derivation needs to be better characterized. Normally, the operational retrieval algorithms are developed based on an existing dataset and are supposed to operate on the upcoming measurements. Therefore, machine learning-based models are supposed to generalize well on the unseen data in future periods. Herein, we aim at the characterization of deep learning capabilities for these GNSS-R operational purposes.</p><p>In this interdisciplinary study, we present a deep learning algorithm processing the CyGNSS measurements to derive wind speed data. The model is supposed to meet an acceptable level of generalization on the upcoming unseen data, and alternatively can be used as an operational processing algorithm. We propose a deep model based on convolutional and fully connected layers processing the DDMs besides ancillary input features. The model leads to the so-far best quality of global wind speed estimates using GNSS-R measurements with a general root mean square error of 1.3 m/s over unseen data in a time span different from that of the training data.</p>

scholarly journals Multimodal Deep Learning for Heterogeneous GNSS-R Data Fusion and Ocean Wind Speed Retrieval

2020 ◽  
Vol 13 ◽  
pp. 5971-5981
Author(s):  
Xiaohan Chu ◽  
Jie He ◽  
Hongqing Song ◽  
Yue Qi ◽  
Yueqiang Sun ◽  
...  
Keyword(s):  
Deep Learning ◽  
Wind Speed ◽  
Data Fusion ◽  
Ocean Wind

NOC GNSS-R Global Ocean Wind Speed and Sea-Ice Products Using Data from the TechDemoSat-1 Mission

2020 ◽  
Author(s):  
Giuseppe Foti ◽  
Matthew Lee Hammond ◽  
Christine Gommenginger ◽  
Meric Srokosz ◽  
Martin Unwin ◽  
...  
Keyword(s):  
Wind Speed ◽  
Sea Ice ◽  
Global Ocean ◽  
Using Data ◽  
Ocean Wind

scholarly journals Temporal variability of GNSS-Reflectometry ocean wind speed retrieval performance during the UK TechDemoSat-1 mission

2020 ◽  
Vol 242 ◽  
pp. 111744
Author(s):  
Matthew Lee Hammond ◽  
Giuseppe Foti ◽  
Christine Gommenginger ◽  
Meric Srokosz
Keyword(s):  
Wind Speed ◽  
Temporal Variability ◽  
Retrieval Performance ◽  
Gnss Reflectometry ◽  
The Uk ◽  
Ocean Wind

scholarly journals Application of Rough and Fuzzy Set Theory for Prediction of Stochastic Wind Speed Data Using Long Short-Term Memory

Atmosphere ◽  
2021 ◽  
Vol 12 (7) ◽  
pp. 924
Author(s):  
Moslem Imani ◽  
Hoda Fakour ◽  
Wen-Hau Lan ◽  
Huan-Chin Kao ◽  
Chi Ming Lee ◽  
...  
Keyword(s):  
Deep Learning ◽  
Wind Speed ◽  
Set Theory ◽  
Rough Set ◽  
Fuzzy Set Theory ◽  
Short Term Memory ◽  
Learning Model ◽  
Short Term ◽  
Long Short Term Memory ◽  
Deep Learning Model

Despite the great significance of precisely forecasting the wind speed for development of the new and clean energy technology and stable grid operators, the stochasticity of wind speed makes the prediction a complex and challenging task. For improving the security and economic performance of power grids, accurate short-term wind power forecasting is crucial. In this paper, a deep learning model (Long Short-term Memory (LSTM)) has been proposed for wind speed prediction. Knowing that wind speed time series is nonlinear stochastic, the mutual information (MI) approach was used to find the best subset from the data by maximizing the joint MI between subset and target output. To enhance the accuracy and reduce input characteristics and data uncertainties, rough set and interval type-2 fuzzy set theory are combined in the proposed deep learning model. Wind speed data from an international airport station in the southern coast of Iran Bandar-Abbas City was used as the original input dataset for the optimized deep learning model. Based on the statistical results, the rough set LSTM (RST-LSTM) model showed better prediction accuracy than fuzzy and original LSTM, as well as traditional neural networks, with the lowest error for training and testing datasets in different time horizons. The suggested model can support the optimization of the control approach and the smooth procedure of power system. The results confirm the superior capabilities of deep learning techniques for wind speed forecasting, which could also inspire new applications in meteorology assessment.

scholarly journals Using Eulerian and Lagrangian Approaches to Investigate Wind-Driven Changes in the Southern Ocean Abyssal Circulation

2013 ◽  
Vol 44 (2) ◽  
pp. 662-675 ◽  
Author(s):  
Paul Spence ◽  
Erik van Sebille ◽  
Oleg A. Saenko ◽  
Matthew H. England
Keyword(s):  
Southern Ocean ◽  
Wind Stress ◽  
Climate Model ◽  
Meridional Overturning Circulation ◽  
Global Ocean ◽  
Lagrangian Particle ◽  
Stress Changes ◽  
Abyssal Circulation ◽  
Flow Pathways ◽  
Ocean Wind

Abstract This study uses a global ocean eddy-permitting climate model to explore the export of abyssal water from the Southern Ocean and its sensitivity to projected twenty-first-century poleward-intensifying Southern Ocean wind stress. The abyssal flow pathways and transport are investigated using a combination of Lagrangian and Eulerian techniques. In an Eulerian format, the equator- and poleward flows within similar abyssal density classes are increased by the wind stress changes, making it difficult to explicitly diagnose changes in the abyssal export in a meridional overturning circulation framework. Lagrangian particle analyses are used to identify the major export pathways of Southern Ocean abyssal waters and reveal an increase in the number of particles exported to the subtropics from source regions around Antarctica in response to the wind forcing. Both the Lagrangian particle and Eulerian analyses identify transients as playing a key role in the abyssal export of water from the Southern Ocean. Wind-driven modifications to the potential energy component of the vorticity balance in the abyss are also found to impact the Southern Ocean barotropic circulation.

Sign in / Sign up

Export Citation Format

Share Document