The global electric circuit land–ocean response to the El Niño—Southern Oscillation

Highlights

• Simulations reveal land–ocean contrast in global electric circuit response to ENSO.

• Contribution of oceans to the global circuit is positively correlated with ENSO cycle.

• Contribution of land areas is, by contrast, negatively correlated with ENSO cycle.

• When summing contributions over all Earth, there is a smaller positive effect of ENSO.

• Key changes in convection are over the Pacific, Maritime Continent and South America.

Abstract

It is known that the global electric circuit (GEC) intensity can be characterised by a single global index, namely the ionospheric potential (IP), made up of contributions from electrified clouds all over the globe. Using the Weather Research and Forecasting model, we have reproduced the atmospheric dynamics for 2008–2018 and simulated the variation of the GEC by parameterising regional contributions to the IP in terms of convection and precipitation. Considering that the El Niño—Southern Oscillation (ENSO) can be quantitatively characterised by sea surface temperatures (SSTs) in the Niño 3.4 region, this allowed us to identify and study in detail the effect of ENSO on regional contributions to the GEC. Our simulations have shown that contributions to the GEC from the land and oceanic parts of the Earth's surface respond oppositely to the ENSO cycle. The oceanic contribution is positively correlated with the Niño 3.4 SST, largely owing to increases in convection over the Pacific Ocean. In contrast to the oceans, the land contribution shows a negative correlation with ENSO due to decreases in convection over the Maritime Continent and South America. The observed correlations are statistically significant and are clearly seen on the decadal timescale; at the same time contributions to the IP for individual years do not always clearly reflect the corresponding Niño 3.4 SST anomalies. During the two El Niños and two La Niñas that occurred between 2008 and 2018, the oceanic contribution always changed in phase with ENSO, increasing in El Niño years and decreasing in La Niña years; on the other hand, the contribution of land showed a clear variation in antiphase with ENSO only for the 2015/16 El Niño and 2010/11 La Niña, characterised by extremely large SST anomalies, with a small and indefinite effect for the two lesser events. When summing the contributions of land and ocean, two strong effects of opposite signs nearly counterbalance each other and we obtain a much less pronounced effect of ENSO on the total IP. This effect is generally positive since land and ocean provide nearly equal contributions to the GEC during Northern Hemisphere winters and, according to our analysis, the contribution of ocean is slightly more sensitive to ENSO than that of land.

Introduction

The El Niño—Southern Oscillation (ENSO) is the largest cause of climate variations on Earth after the seasonal cycle of summer and winter. The ENSO is made up of irregular cycles, each of which contains a warm phase (El Niño) and a cold phase (La Niña), driven by changes in sea surface temperatures (SSTs) in the central and eastern equatorial Pacific Ocean (see, e.g., McPhaden et al., 2020). During ENSO events the SSTs in the eastern Pacific increase (in the case of El Niños) or decrease (in the case of La Niñas) by a few degrees for extended periods of 12–18 months. These SST changes occur over very large areas for extended periods of time and hence have a direct influence on the global climate system with major impacts on global temperature patterns (Tsonis et al., 2005) and global rainfall patterns (Villafuerte II and Matsumoto, 2015).

It is thought that most of these changes are related to shifts in the atmospheric convection patterns. Both the north—south Hadley circulation and the east—west Walker circulation shift during ENSO events, with implications on the hydrological cycle around the globe. These circulation patterns are related to tropical deep convection, which is often associated with thunderstorms and lightning activity. Therefore it is clear that the ENSO cycle has a substantial influence on the Earth's electrical environment; this influence has been discussed in a number of studies, whose review can be found in Williams and Mareev (2014, section 9). Nearly all of these studies investigated the impact of ENSO on lightning activity (e.g., Sátori and Zieger, 1999; Goodman et al., 2000; Hamid et al., 2001; Price and Federmesser, 2006; Sátori et al., 2009; Price, 2009). However, such investigations were generally focused on specific regions and no global analysis has been performed; this is due to the lack of global data related to thunderstorms and electrified clouds on long timescales needed to study ENSO variability.

Despite a considerable amount of research devoted to investigating the effect of ENSO on lightning, little attention has been given to its impact on quasi-stationary fields and currents in the atmosphere. At the same time it is known that such fields and currents are part of the so-called global electric circuit (GEC; see, e.g., Rycroft et al., 2000; Tinsley, 2008), which is a natural concept to consider when analysing global effects of climate variations on atmospheric electricity. The idea of the GEC is based on the hypothesis proposed by Wilson, 1921, Wilson, 1924 according to which the charging of thunderstorms and electrified shower clouds (ESCs) maintains the distribution of quasi-stationary fields and currents in the entire atmosphere. More precisely, it is thought that charge separation inside electrified clouds (both thunderstorms and ESCs) drives the electric current upwards to the ionosphere (Mach et al., 2010, Mach et al., 2011), and this current eventually returns to the Earth's surface in fair-weather regions (e.g., deserts). The highly conducting (hence nearly equipotential) Earth and ionosphere complete the electrical network; the potential difference between them is called the ionospheric potential (IP; Markson, 1986, Markson, 2007).

Thus the GEC links together thundestorm regions and fair-weather regions. Experimental evidence supporting this concept includes the well-known observation that the diurnal variation of the fair-weather electric field at the Earth's surface in universal time (UTC) looks nearly the same at different locations around the globe (this dependence is called the Carnegie curve; see Harrison, 2013) and is tightly correlated with the diurnal cycle of global thunderstorm activity (Whipple and Scrase, 1936). Another observation corroborating the concept of the GEC is the fact that simultaneous measurements of the IP (i.e. of the potential of the lower ionosphere relative to the Earth's surface) performed at remote locations yield similar results (Markson et al., 1999), which suggests that the lower ionosphere indeed has the same potential all over the globe, maintained by global thunderstorms and ESCs.

Until recently, the only known study aimed at investigating the effect of ENSO on the GEC was the one by Harrison et al. (2011), who analysed the results of atmospheric electricity measurements at the Shetland Islands during 1968–1984. They found a positive correlation between the mean fair-weather potential gradient in December and the ENSO cycle and identified the hours when the potential gradient is most sensitive to ENSO variability. However, atmospheric electricity measurements are not able to reveal the mechanisms behind the observed links between ENSO and the GEC, as the measured potential gradient values are determined by the sum of contributions to the GEC from electrified clouds all over the globe and it is impossible to determine which regions contributed the most to this value.

In this regard, it is particularly helpful to employ theoretical modelling. Global models of atmospheric dynamics can both provide ENSO indices of SST and surface pressure anomalies (e.g., Southern Oscillation Index) and simulate the convection in response to changes in these SSTs. In addition, recent studies (Slyunyaev et al., 2019; Ilin et al., 2020) have developed a methodology of estimating the spatial and temporal distribution of electrified clouds in these models and their contributions to the GEC. Recently Slyunyaev et al. (2021) have employed this approach to identify the influence of ENSO on the shape of the diurnal variation of the GEC; in particular, they have shown that pronounced anomalies in the relative IP around 13:00 UTC and 21:00 UTC precisely correspond to strong El Niño and La Niña events.

In this study we present another link between ENSO variability and changes in the IP (which characterises the GEC intensity). Instead of regarding the IP as a single integrated global quantity, we focus on contributions to the total IP from different regions of the globe. Using the Weather Research and Forecasting model (WRF; see Skamarock et al., 2008), we reproduce the global atmospheric dynamics on the basis of real meteorological data over a decade covering several ENSO events. Then a parameterisation of the IP allows us to investigate the impact of ENSO on regional contributions to the GEC and to reveal the cumulative effect of all such regional links, thereby elucidating the relationship between ENSO and the GEC.