Long‐Term Estimation of Diurnal Vertical E × B Drift Velocities Using C/NOFS and Ground‐Based Magnetometer Observations

Abstract

We report on the development of a new mathematical expression to estimate local daytime (0700–1700 LT) vertical E × B drift in low latitudes using a combination of ground‐based magnetometer measurements and Communications and Navigation Outage Forecasting System (C/NOFS) satellite observations. The expression was developed over Jicamarca (11.8°S, 77.2°W; 0.8°N geomagnetic) and validated with Jicamarca Unattended Long‐Term studies of the Ionosphere and Atmosphere (JULIA) mode and incoherent scatter radar (ISR) measurements during the period 2008–2014. The obtained correlation coefficient (R) values computed using observed and derived vertical E × B drift velocities are 0.79 and 0.84 for ISR and JULIA, respectively when data are available during 2008–2014. Storm‐time comparison between observed and derived vertical E × B drift velocities agreed well with R of 0.92 and 0.87 during 5–8 August 2011 and 8–11 March 2012 geomagnetic storm periods for ISR and JULIA observations, respectively. Overall, we found that the developed expression is applicable in estimating vertical E × B drift response during quiet and geomagnetic storm periods. Based on these findings, we suggest that it is possible to develop accurate daytime global vertical E × B drift model over the equatorial latitude regions using inexpensive magnetometer observations and available satellite data.

Counter‐Electrojet Occurrence as Observed From C/NOFS Satellite and Ground‐Based Magnetometer Data Over the African and American Sectors

“…The derived Δ H data from PUER‐LETI pair of magnetometer stations is mostly available from 2009 and further contains significant data gaps especially in 2010–2011, explaining the difference in data points displayed in scatter plots of Figure for PUER and AAE. The correlation values are comparable to earlier results which reported values of 0.57 and 0.51 over Jicamarca (Habarulema et al, ) and AAE (Dubazane & Habarulema, ), respectively, between C/NOFS vertical E × B drift and Δ H . The low correlation values are attributed to the altitude differences at which C/NOFS vertical E × B drift and Δ H are computed.…”

“…Due to the unusual extended solar minimum at the end of solar cycle 23 during 2008–2010 (e.g., Chen et al, ; Ezquer et al, ) when vertical E × B drift did not show expected direct relationship with solar activity (e.g., Dubazane & Habarulema, ; Habarulema et al, ), the presentation of results is categorized into two periods of 2008–2010 and 2011–2014, respectively, for both the American and African sectors. This was however done only for the 69°W (American) and 38°E (African) longitude sectors where 2008–2014 data were available; otherwise, 2011–2014 data sets were analyzed for the 56°W (American) and 9°E (African) longitude sectors.…”

“…Due to the intention of directly comparing with magnetometer‐derived EEJ, it was necessary to limit the latitudinal coverage of satellite's data around the geomagnetic equator by constraining the latitude of observations to remain within the EEJ band. For this purpose, the latitude range used was ±4° around the geomagnetic equator and within a longitude range of ±8° centered around the meridian of the pair of magnetometer stations (e.g., Dubazane & Habarulema, ; Habarulema et al, ) for both African and American sectors. In an effort to minimize potential outliers associated with C/NOFS vertical ion plasma drift data within 400–550 km, we employed the median and scaled median absolute deviation (e.g., Huber, ; Huber & Ronchetti, ) for each satellite track during the entire period (2008–2014) of analysis.…”

“…This procedure is based on differencing the horizontal component measurements (after removing the average nightime baseline value) of a magnetometer displaced by 6–9° from the geomagnetic equator from the corresponding H values measured by the magnetometer at the equator. The estimated EEJ is a proxy of low latitude vertical E × B drift (e.g., Anderson et al, ; Habarulema et al, ; Yizengaw et al, ). Figure shows local daytime changes in H component after removing the nightside (2300–0300 local time) baseline measured at the equator and off the equator and the differences between the two representing the EEJ for randomly selected days over the American (PUER‐LETI [23 November 2011] and ALTA‐CUIB [28 October 2012]) and African (ABJA‐CMRN [23 January 2012] and AAE‐ETHI [19 November 2008]) regions.…”

“…In an effort to minimize potential outliers associated with C/NOFS vertical ion plasma drift data within 400-550 km, we employed the median and scaled median absolute deviation (e.g., Huber, 1981;Huber & Ronchetti, 2009) for each satellite track during the entire period (2008-2014) of analysis. This filtering method has been previously used in related analyses (e.g., Habarulema et al, 2018;Lomidze et al, 2017) and is demonstrated in Dubazane and Habarulema (2018) for C/NOFS satellite data treatment. After removing outliers, the C/NOFS data are averaged in 3-min intervals.…”