Automatic Detection of Interplanetary Coronal Mass Ejections from In Situ Data: A Deep Learning Approach
Gautier Nguyen, Nicolas Aunai, Dominique Fontaine, Erwan Le Pennec, Joris Van den Bossche, Alexis Jeandet, Brice Bakkali, Louis Vignoli, and Bruno Regaldo-Saint Blancard The Astrophysical Journal, Volume 874, Number 2 doi : https://doi.org/10.3847/1538-4357/ab0d24
Decades of studies have suggested several criteria to detect interplanetary coronal mass ejections (ICME) in time
series from in situ spacecraft measurements. Among them, the most common are an enhanced and smoothly
rotating magnetic field, a low proton temperature, and a low plasma beta. However, these features are not all
observed for each ICME due to their strong variability. Visual detection is time-consuming and biased by the
observer interpretation, leading to non-exhaustive, subjective, and thus hardly reproducible catalogs. Using
convolutional neural networks on sliding windows and peak detection, we provide a fast, automatic, and multiscale
detection of ICMEs. The method has been tested on the in situ data from WIND between 1997 and 2015, and
on the 657 ICMEs that were recorded during this period. The method offers an unambiguous visual proxy of
ICMEs that gives an interpretation of the data similar to what an expert observer would give. We found at a
maximum 197 of the 232 ICMEs of the 2010–2015 period (recall 84%±4.5%), including 90% of the ICMEs
present in the lists of Nieves-Chinchilla et al. and Chi et al. The minimal number of False Positives was 25 out of 158 predicted ICMEs (precision 84%±2.6%). Although less accurate, the method also works with one or several
missing input parameters. The method has the advantage of improving its performance by just increasing the
amount of input data. The generality of the method paves the way for automatic detection of many different event
signatures in spacecraft in situ measurements.
SPACEOBS : a space incubator funded by University Paris-Saclay started this January. The first part of the project is for 2017. Funding is 700k€.
SPACEOBS will develop inovative activities in four essential directions:
- MultiScale multi-spacecraft data analysis (WP1)
- Coupling simulations with observations (WP2)
- New generation instrument design (WP3)
- Public outreach and teaching (WP4)
I’m co-leading the work package 1
Cold ion heating at the dayside magnetopause during magnetic reconnection
Geophysical Research Letters, Volume 43, Issue 1, pp. 58-66 DOI : http://dx.doi.org/10.1002/2015GL067187
Toledo-Redondo, S.; André, M.; Vaivads, A.; Khotyaintsev, Yu. V.; Lavraud, B.; Graham, D. B.; Divin, A.; Aunai, N.
Cold ions of ionospheric origin are known to be present in the magnetospheric side of the Earth’s magnetopause. They can be very abundant, with densities up to 100 cm-3. These cold ions can mass load the magnetosphere, changing global parameters of magnetic reconnection, like the Alfvén speed or the reconnection rate. In addition they introduce a new length scale related to their gyroradius and kinetic effects which must be accounted for. We report in situ observations of cold ion heating in the separatrix owing to time and space fluctuations of the electric field. When this occurs, the cold ions are preheated before crossing the Hall electric field barrier. However, when this mechanism is not present cold ions can be observed well inside the reconnection exhaust. Our observations suggest that the perpendicular cold ion heating is stronger close to the X line owing to waves and electric field gradients linked to the reconnection process.
BV technique for investigating 1-D interfaces
Published in Journal of Geophysical Research
Dorville Nicolas, Belmont Gerard, Rezeau Laurence, Aunai Nicolas Retinò, Alessandro
To investigate the internal structure of the magnetopause with spacecraft data, it is crucial to be able to determine its normal direction and to convert the measured time series into spatial profiles. We propose here a new single-spacecraft method, called the BV method, to reach these two objectives. Its name indicates that the method uses a combination of the magnetic field (B) and velocity (V) data. The method is tested on simulation and on Cluster data, and a short overview of the possible products is given. We discuss its assumptions and show that it can bring a valuable improvement with respect to previous methods.