Report on WP5 – Deliverable 5.4.1
IZMIRAN
Hourly CR variations and their pitch angle and
longitude distributions
E. Eroshenko, A. Belov, A. Asipenka, V. Yanke
IZMIRAN, June 30, 2009
1
Document information
Project
Project acronym:
Project full title:
Grant agreement no:
Funding scheme:
Project start date:
Project duration:
Call topic:
NMDB
Real-time database for high
resolution Neutron Monitor
measurements
213007
Combination of Collaborative
Projects & Coordination and
Support Actions
01.01.2008
24 months
INFRA-2007-1.2.1 Scientific Digital
Repositories
Project web-site:
www.nmdb.eu
Deliverable Report: WP5.4.1
Period Covered:
Actual Submission Date:
Editors:
Authors:
IZMIRAN
01.01.08 – 30.06.2009
30.06.2009
E. Eroshenko, V. Yanke
A. Belov, A. Asipenka, E.
Eroshenko, V. Yanke
Final
10
Real time longitudinal distribution,
precursors
Draft/Final :
No of pages (including cover)
Keywords:
2
Contents
1. Introduction
2.
3
………… 4
Data and Station Ring method
2.1. NMDB data usage……………………………………………………….5
2.1.1. Neutron monitor data downloading
…. …6
2.2. Other sources data usage
....6
3. Description of the site of internet project
6
3.1. The CR variation distribution by asymptotic longitudes and examples of
precursors by retrospective data
7
3.2. Asymptotic distribution by real time data from NMDB
8
4. Summary
9
5.4.1 Hourly CR variations and their pitch angle and longitude
distributions (IZMIRAN, NKUA, Kosice, TAU) Month 18
This part is very important for searching for the precursors of Forbush effects and
geomagnetic storms by the data of world wide NM network, using the method of
“station ring”.
1. INTRODUCTION.
The galactic cosmic rays (GCR) interact with transients moving in the space toward the
Earth. Due to their high speed they might provide the information about coming disturbed
region well advance its arriving to Earth. Existence of so-called ‘precursors’ in the CR
registered by ground detectors, was supposed almost from the very beginning of continuous
CR registration by a network of stations. Later it has been detected in the data of neutron
monitors before the beginning of strong magnetic storms and large Forbush effects, and since
that time a detailed researches of those effects started. Precursor effect consists of
combination of two kinds of galactic CR variations: pre-decrease and pre-increase, on a
miscellaneous distributed by various CR stations at the same time before the shock arrival.
Precursory decreases apparently result from a “loss-cone” effect, in which a neutron monitor
station is magnetically connected to the cosmic ray-depleted region downstream the shock
(Fig.1). Pre-increase is usually caused by particles reflecting from the approaching shock.
The many of predictors have a peculiar longitude (or, pitch-angle) dependence of CR
intensity with the abrupt transfer from minimum to maximum of CR variations which cannot
be fitted by the sum of only the first two harmonics.
3
Intensity deficit
confined in a cone
Figure 1. Illustration of “loss cone” effect in cosmic ray variations observable at Earth.
These sharp transfers occur most probably within the 140-180and 270310longitude regions, near the usual direction of the magnetic field line (sunwards and
anti-sunward). These anomalies are most often observed in the last hours before shock
arrival, typically within 4 hours. However, they sometimes span a longer period, and persist
downstream of the shock. The NMN is a good tool for detecting such anomalies in the pitchangle or longitudinal distribution. At present, when data of many stations are accessible in
real time, it would be desirable to search such anomalies in real time mode and use this
information in the short time forecasting of geomagnetic activity.
Precursory effect is very anisotropic, thus the sky coverage in the asymptotic
directions of the stations used, should be as full as possible, otherwise we can miss any
precursor. In Fig. 2, 2a the longitudinal distributions of the observed CR variations in the
event on September 1992 are presented by the full number of neutron monitor network and
calculated for a limited numbers as well. It is seen that we lose very strong precursor if the
number of stations is limited within very narrow longitude range.
In the Bartol Research Institute the system of real time monitoring of the precursors is
created
on
the
base
of
pitch
angle
distribution
http://neutronm.bartol.udel.edu//spaceweather/. They use a limited number of stations (only
9) but distributed homogeneously by longitudes (Spaceship Earth). The system is working
well but it depends on the reliability of all station operating (break at one station create a gap
in the scanning of ~500 of celestial sphere) and on a persistence of interplanetary magnetic
field (IMF) data which are used for pitch angle calculations.
IZMIRAN offers another approach based on the distribution of CR variation by
asymptotic longitudes. It is defined from the ‘ring station method’, and this procedure is
more easy and reliable, and not less informative. Creation of NMDB in the frame of EC FP7
project “Real time neutron monitor database with short time resolution” allows the use
hourly data from many stations (about 23 at present) for constructing and plotting asymptotic
longitudinal distribution of the CR variations at any time. This application is realized in the
IZMIRAN and located by the addresses http://cr0.izmiran.ru/PrecurcorMonitoring/index.htm
and http://www.nmdb.eu/?q=node/19, PRECUROR MONITORING so far.
Four Teams participated in developing of this Application. The main contribution is
made by IZMIRAN. TAU did help with elaboration of the algorithm fitted to real time data.
Kosice contributed with a creation of web page of Internet project. NKUA tested the
software and using the results for analysis on the retrospective data.
4
2. DATA AND STATION RING METHOD
For searching and study of precursory effects hourly data from NM network are
usually used in the Ring Station (RS) method. It is preferable to use stations with cutoff
rigidity Rc<4 GV and homogeneously distributed by the ring around the globe (Station Ring
method), but we elaborated the simplified version specially for real time data allowing the
use also of stations with higher rigidities (Rome, Athens, Almaty, Jungfraujoch) from
NMDB.
Shift of the effective position of station from its geographic longitude because of the Earth
magnetic field was taken into account by using the coupling coefficients [6] for the flat
rigidity spectrum of the first harmonic up to 100 GV. Thus, at each hour we can plot a
distribution of CR variations by the asymptotic longitudes. This method provides obtaining
of different than harmonic longitudinal distribution of CR intensity and it is less depended on
an uncertainty of the model of isotropic and anisotropic variations.
Fig. 2. Distribution of CR variations by the asymptotic longitudes of different stations (black circlesdecrease, grey- increase of CR variations) before the shock arrival (vertical line) in September 1992.
This picture is plotted by 45 world wide network stations and shows clearly pronounced pre-decrease
at the stations viewing along IMF force line (around 1350).
Fig. 2ab. The same event but plotted only by European stations (left side) and by European + Russian
NMS (right side).
Precursory effect is very anisotropic, thus the sky coverage in the asymptotic directions of
the stations, should be as full as possible, otherwise we can miss any precursor. In Fig. 2 the
longitudinal distributions of the observed CR variations in the event on September 1992 are
plotted by the full number of neutron monitors, and by the limited numbers as well. It is seen
5
that we lose very strong precursor if number of stations is ranged within very narrow
longitude belt.
2.1. NMDB data usage
This fact makes it incumbent to use data from the stations distributed by the globe with
overlapping of the celestial sphere by the asymptotic directions. We included for use hourly
data from NMDB for the stations: Almaty, Apatity, Athens, Baksan, NorAmberd, Aragats,
ESOI, Irkutsk, Irkutsk2, Jungfraujoch (IGY and 3NM64), Kerguelen, Kiel, Lomnitsky Stit,
Magadan, Moscow, Mirny, Novosibirsk, Oulu, Rome, Terre Adelie, Tixie Bay, Yakutsk.
Some of them still are not ready for continues quality data presenting (NorAmberd, Aragats,
Esoi, Kiel, Novosibirsk), but we hope, this is a question of the nearest future.
2.1.1. Neutron monitor data downloading
Special scripts were elaborated to get real time hourly data from NMDB. The
procedure of data downloading from the NMDB is present by the address:
http://cosmicrays.oulu.fi/nmdbinfo/?q=node/35 in the attached file ‘Read-writeIZMIRAN.pdf’. But we should remember that real time data at some stations have not been
subjected to rigorous quality control; it may contain "glitches" that produce false alarms or
fail to detect true space weather disturbances.
2.2. Other sources data usage
Neutron monitor hourly data from the western stations (Forth Smith, Pewanyuk, Newark,
McMurdo, Thule) are also used (from USA, Bartol group: http://neutronm.bartol.udel.edu/)
to complete scanning of celestial sphere. All data are retrieved into the local DB
(http://cr0.izmiran.rssi.ru/common/All_CR_stations_H.htm) which is implicated in different
Applications.
There are also relevant data on the solar wind and geomagnetic activity
(http://cr20.izmiran.ru/sw/main.htm) which would be necessary for further analysis.
3. Description of the site of internet project
The interface of Internet project for plotting a distribution of hourly CR variations at any
moment
0
http://cr0.izmiran.ru/PrecurcorMonitoring/index.htm),
or,
http://www.nmdb.eu/?q=node/19, PRECUROR MONITORING, is shown in Fig. 3.
Some explanations to this site:
“Monitoring (NM Network) Precursors Retro and Real Time”
This line opens a picture “Monitoring of CR anisotropy”, where one can see asymptotic
longitudinal distribution of CR variations for the current hour. It is possible to get such a
distribution for any time interval selecting it and base period in the table ‘PARAMETERS”
on this site.
“Data for precursor analysis for last period” Digital Plot Data Plot Variations
Clicking this line you can look data from all stations (separately and altogether) used at a
current moment in digital and graphic forms that gives a chance for brief control of data
quality at different stations.
6
Fig.3. Presentation of the software for plotting hourly CR variations by the asymptotic
longitudes and demonstration of the magnetic storm precursors.
There are also the lines to access the Bartol Space Weather site (for comparing) and relevant
data on the solar and geomagnetic activity (for analysis in the case).
3.1. Longitudinal distribution of the CR variations obtained by retrospective
data and examples of precursors
Elaborated software can provide also graphic presentation of longitudinal distribution
of the CR variations for any selected period by the retrospective data which also can be taken
from NMDB. Below some examples are presented which demonstrate features of application
and
possibility
of
a
method
of
ring
stations.
Fig. 4. Asymptotic longitude distribution of CR variations before the Forbush effect
(FE) and magnetic storm (SSC-vertical line) in August 1979. Size of circles correspond to
the magnitude of CR variation relatively to quiet period: yellow- increase, red-decrease.
In this case a loss cone effect is well indicated in the longitudinal-time distribution of
the CR variations. One can see a decrease within the narrow longitudinal sector along the
7
average direction of the IMF (~1300 ) during ~11 hours before the shock. This peculiar
distribution is evidently a precursor of a large FE, which most probably was much greater to
the west of Sun-Earth line than that observed near Earth. This is an evidence of the strong
magnetic fields in the solar wind disturbance which the Earth entered that time, and caused
by this a possibility of the strong magnetic storm evolving. In that moment Kp index was 5-,
but in 10 hours it really reached the level 8 (severe magnetic storm).
Fig. 5. Well know GLE occurred on June 15, 1991 on the background of the strong
geomagnetic storm. Vertical lines mean SSC (moment of shock arrival), circles correspond
to the increased (yellow) and decreased (red) CR variations.
In this figure the well known period in the mid of June 1991 is shown, when in the
neutron monitor data both an increase of solar CR and very large decrease of galactic CR
were observed. However, there is a quiet period also in the very beginning (0-5 hours) where
usual first harmonic of CR anisotropy is seen in a longitudinal distribution.
Fig.6. Distribution of the CR variations by the asymptotic longitudes obtained from
the neutron monitor network data by the ring station method for the event in December 2006.
8
There is a series of FEs in December 2006 started after arriving of the interplanetary shock
and SSC on December 8. On this background the enhancement of solar cosmic rays (GLE)
on December 13 stands out. One can see that the onset of GLE at the NMs collected CR
from the western directions, started earlier and was more significant than those recorded
within other longitudinal zones.
These examples demonstrate some properties of longitude-time distribution of the CR
variations and possibility of this method usage for diagnostics of the interplanetary and near
Earth space.
3.2. Asymptotic distribution by real time data from NMDB
Fig. 7.
When you click “Monitoring (NM Network) Precursors Retro and Real Time” on the
site in fig. 3 illustrated, the picture of longitude-time distribution of CR variation in real time
should appear on the screen. This distribution is plotted for the current hour, and CR
variations are automatically calculated relatively the first hour in current day. But you may
choose any other period for base and for distribution plotting (division ‘Prameters’). You
may look digital CR variations for all implemented stations at considered time clicking
“Digit”.
During last years no one large FE or strong magnetic storm occurred, so we could not
observe any precursor in real time, we see quiet picture of distribution corresponding to
usual daily variation. But we are ready to register the events when disturbances appear.
IV. CONCLUSIONS
Creation of NMDB in the frame of EC FP7 project “Real time neutron monitor
database with short time resolution” allows the use hourly data from many stations (about 23
at present) for constructing and plotting asymptotic longitudinal distribution of the CR
9
variations at any time. This application is realized in IZMIRAN and located by the address
http://cr0.izmiran.ru/PrecurcorMonitoring/index.htm so far.
Using mentioned above events as the examples (actually we considered much more events) it
can be argued that majority of FDs have the predictors revealed itself by different way in the
CR behaviour some hours before the FD onset. The ”station ring” method gives a good
possibility to detect and study these predictors by NMs data and investigate the structure of
transients and character of their moving. Another kind of studies which has never been
carried out yet is the structure and propagating transient from the coronal holes. Now, with a
creation of real time nmdb database there is a chance to make precursor monitoring in real
time and investigate both sporadic and recurrent solar-terrestrial events.
Creation of NMDB in the frame of EC FP7 project “Real time neutron monitor
database with short time resolution” allows the use hourly data from many stations (about 23
at present) for constructing and plotting asymptotic longitudinal distribution of the CR
variations at any time. This application is realized in IZMIRAN and located by the addresses
http://cr0.izmiran.ru/PrecurcorMonitoring/index.htm and http://www.nmdb.eu/?q=node/19,
PRECUROR MONITORING so far.
10