See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/252693909

Correlation between solar activity and Earth's ionospheric electron content

during the 23rd solar cycle

Article · December 2010

CITATIONS READS

3 778

9 authors, including:

Juliette Legrand Robert Burston

Royal Observatory of Belgium University of Bath
68 PUBLICATIONS   1,370 CITATIONS    14 PUBLICATIONS   59 CITATIONS   

SEE PROFILE SEE PROFILE

Carine Bruyninx P. Defraigne

Royal Observatory of Belgium Royal Observatory of Belgium
184 PUBLICATIONS   1,463 CITATIONS    171 PUBLICATIONS   2,589 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

DEMETRA View project

DEMETRA Horizon 2020 View project

All content following this page was uploaded by Juliette Legrand on 02 June 2014.

The user has requested enhancement of the downloaded file.

 Correlation Between Solar Activity and Earth’s Ionospheric Electron
Solar-Terrestrial
Content During the 23 Solar Cycle.
rd
Royal Observatory
Center of Excellence of Belgium
N. Bergeot1, J. Legrand1, R. Burston1, C. Bruyninx1, P. Defraigne1, J.-M. Chevalier1, F. Clette2, C. Marque2, L. Lefevre2
Abstract number : SA33B-1774 nicolas.bergeot@oma.be
1. Reference Systems and Planetology, Royal Observatory of Belgium, Brussels, Belgium.
2. Solar Physics and Space Weather, Royal Observatory of Belgium, Brussels, Belgium.

Introduction 2. Solar Data vs. Ionospheric Data 4. Latitudinal Correlations

One of the future challenges of the Space Weather community will be to predict the Earth’s ionospheric state in The Daily Sunspot Number
response to variations of the Solar activity (Fig. 1). The Earth’s ionosphere is the atmospheric layer degrading the The daily Sunspot Number (SN) provides the longest available record of Solar activity since 1818 and is the main
most radio communications, particularly the GNSS (Global Navigation Satellite Systems, e.g. GPS, GLONASS, reference Solar index. SN represents the recurrence of intense magnetic loops visible on the Sun surface. Presently,
GALILEO). the estimated International SN (Fig. 3) is distributed by the Solar Influences Data Center (SIDC) and is based on
statistical processing of the data of a worldwide network of 86 stations from 29 countries [Clette et al., 2007].
The beginning of the 23rd Solar cycle (May 1996 to December 2008) coincided with the start of the catalogue of
global ionospheric modeling based on GNSS data from dense ground networks. In addition, many parameters of the F10.7 Solar Flux
Solar activity are historically measured.
The integrated emission from the Solar disc
density at 2800 MHz (10.7cm wavelength called
In this study, we compared the daily Sunspot Number (SN) and F10.7 flux with the
F10.7) is recorded routinely [Covington, 1969]
daily mean ionospheric Total Electron Content (TEC) obtained from GNSS data
since 1947.
during the 23rd Solar cycle in order to better understand the ionization response to
Solar activity variations. The main questions we address are : Two sets of fluxes are used here (Fig. 3) :
(1) the observed flux (F10.7o) which is the true
 What are the correlations between Solar and ionospheric states ? measured values,
(2) the adjusted flux (F10.7a) which describes
 Which Solar parameter shows the best correlation with ionospheric TEC ? more purely the Sun's behavior (corrected for
Do the correlations vary in time (seasons, Solar cycle phases) ?
NASA/National Space Science and Technology Center
the changing distance between the Earth and
Figure 1 : Sun activity vs. Earth atmosphere. Sun throughout the year).
 Is there a geographic dependency of the correlations ?
The 10.7cm Solar flux is expressed in Solar Flux
Unit (1 SFU = 10-22 Wm-2Hz-1). In this study we
used the daily F10.7 estimated from the radio
1. Ionospheric Data telescope in Penticton, British Columbia and
accessible via the NGDC-NOAA website.
Global Ionospheric Maps (GIMs)
GIMs from the CODE Analysis Center [Schaer et al., 1998] are used as ionospheric data. VTEC Data
The CODE GIMs product (Fig. 2) consists of global Vertical Total Electron Content (VTEC) estimated from GNSS The daily mean VTEC in TECU (Fig. 3) is
data. The GIMs are modeled in a Solar geomagnetic reference frame using a spherical harmonics expansion up to estimated from the GIMs (see section 1). It
corresponds to the average of the daily VTEC.
degree and order 15 with piecewise linear functions for the temporal evolution of the VTEC. An ionospheric single thin
layer (ISTL) hypothesis is used to express the VTEC at a given altitude (typically 400-450 km, in the F region of the We study the correlation between Solar (SN,
ionosphere where the electron concentration is maximum). F10.7o/a) and ionospheric (VTEC) data from
1995-001 to 2009-275 (5389 days).
3. Daily Mean Global Correlations
To characterize the link between Solar parameters (SN and
In this section, the latitudinal dependence of the correlation between the VTEC and the observed F10.7o is
investigated. The daily mean VTEC is computed each 10° of latitude between N80° to S80° by estimated the mean
daily VTEC from a band of 20° (in latitude) around the given latitude.
Coordinates System Dependency of the Correlation VTEC/F10.7o
In Fig. 5 the VTEC is expressed in geographic and geomagnetic
coordinates in order to test the effect of the coordinate system on the
VTEC/F10.7o correlation. The geomagnetic coordinates are given by the
IGRF11 model [Finlay et al., 2010, see Fig. 1] using a dipole
approximation of the geomagnetic field.
Results 2
 The correlation between mean daily latitudinal VTEC expressed in
geomagnetic coordinates and F10.7o does not significantly change
except close to 30°S of latitude.
 The maximum correlation (R=0.90) is observed close to N10°-
N20°.
 A dichotomy is identified with a lower correlation in the Southern
Figure 5 : Latitudinal dependency of the VTEC/F10.7o
correlation with respect to the coordinates system hemisphere (R=0.81±0.06) compared to the North (R=0.86±0.03).
used to extract the VTEC daily mean.
Seasonal Dependency of the Correlation VTEC/F10.7o
To study the VTEC response to relative position and orientation of the
Earth/Solar system, the latitudinal daily mean VTEC and F10.7o are
divided in four seasonal periods : Spring, Summer, Autumn and Winter
corresponding to seasons in the Northern Hemisphere (Fig. 6).
Results 3
 The maximum correlation (R=0.94) is close to S30° latitude during
winter while the minimum (R=0.74) occurs during summer period
Figure 3 : Ionospheric and Solar data during the 23rd Solar Cycle. Bottom : daily
mean global VTEC in TECU. Middle : Sunspot Number as delivered by SIDC. Top : close to the South polar .
F10.7 flux observed and adjusted from Penticton radio telescope.
 The correlation is improved during Autumn (R=0.89±0.02)
Figure 6 : Latitudinal dependency of the VTEC/F10.7o compared to Spring and Summer periods (R= 0.87±0.04).
correlation with respect to the North Hemisphere
seasons. DOY= Day Of Year.
Solar Cycle Dependency of the Correlation VTEC/F10.7o
N

F10.7) and the ionospheric VTEC, we estimated the linear  xy  ( x  x).( y

i i  y)
The latitudinal daily mean VTEC and the F10.7o are now divided in three
R  i 1
correlation coefficient R between SN/VTEC and F10.7/VTEC  x y N N Solar cycle phases (Fig 7) :
from the following equation :  ( xi  x) .
i 1
2
 i
( y
i 1
 y ) 2
oSolar minimum activity phases : 1995-1997 and 2007-2009.7
o Solar transition activity phases : 1997-2000 and 2002.5-2007
R is estimated (Fig. 4) for 0-day, 1-day and 2-days time offsets in order to quantify if the correlation increases when o Solar maximum activity phase : 2000-2002.5
considering a delay in the ionospheric response to Solar activity.
Results 4
 The correlations are consistent for the two transition and minimum
Figure 2 : Global VTEC map in TECU. GIM map from CODE at 12:00 UT, June 30, 2010 or Day Of Year phases.
(DOY) 181. The black triangles represent the GNSS stations (253 sites in 2010) used to model the  The correlation is larger during Solar transition activity phases
ionosphere. White lines are the geomagnetic coordinates (dipole model) from IGRF11 [Finlay et al., 2010].
(R=0.71±0.09) and decreases during maximum (R=0.49±0.05) and
GIMs are provided as daily files containing global VTEC estimated each 2 hours on a 2.5°x5° (latitude/longitude) grid minimum (R=0.32±0.19) phases with a complete de-correlation at
since 1995. Presently, more than 250 GNSS ground stations data are used to generate such models. VTEC GIMs Figure 7 : Latitudinal dependency of the VTEC/F10.7o
N50° during Solar minimum.
are expressed in TEC Units (1 TECU=1016 e-/m2) with RMS between 0.7 to 6 TECU and accuracy of the order of 2 to correlation with respect to the Solar Cycle Phases.
8 TECU.
Conclusion
Since 1995, the GIMs characteristics vary with the evolution of the ground GNSS network, satellite constellations and
data processing strategy (Tab. 1). From the study of the correlations between ionospheric VTEC extracted from CODE GIMs and Solar indexes we can
draw two major conclusions:
Figure 4 : Daily Mean global VTEC vs. daily Solar parameters during the 23rd Solar Cycle. The figures show the typical data sets used to estimate
Period (yyyy-ddd) Time span Altitude (km) of the # GNSS Stations # GNSS satellites the linear correlation coefficient between daily mean global VTEC and a) daily Sunspot Number, b) daily F10.7 adjusted and c) daily F10.7 observed. For  The correlation is maximum when :
VTEC maps type each parameter we considered a time delay between Solar and ionospheric parameters : Blue = 0-days; Red = 1-day and Green = 2-days. R is the
Beginning End (year) ISTL Min.-Max. Min.-Max
estimated linear correlation coefficient. Considering F10.7 observed vs. VTEC.
1995-001 1997-032 2.09 400.0 1 map at 12:00 UT 50-86 ~24 ?
1997-033 1997-054 0.06 450.0 Each 2 hours (12 maps) + RMS 72-92 24-25 During winter period in the Southern Hemisphere.
1997-055 1998-086 0.09 400.0 1 map at 12:00 UT 42-99 24-27 Results 1 During Solar transition phase activity.
1998-087 1998-098 0.03 450.0 Each 2 hours (12 maps) 76-80 27-27
 The best correlation (R=0.91) is obtained when considering the observed F10.7o. When geomagnetic coordinates are considered for VTEC extraction.
1998-099 2009-275 12.48 450.0 Each 2 hours (12 maps) + RMS 73-261 25-36
 The correlation does not change when increasing the time offset (0-2 days) between daily mean global  There is a clear dichotomy between the Southern and Northern Hemisphere concerning the response of the
Table 1 : Evolution of the CODE GIMs characteristic used in this study. VTEC and Solar parameters. ionosphere to Solar activity.
The authors acknowledge the CODE IGS Analysis Centre for their GIM maps (ftp://ftp.unibe.ch/aiub/CODE) as well as NGDC_NOAA for the F10.7 data (ftp://ftp.ngdc.noaa.gov/STP/SOLAR_DATA/)

View publication stats