The search for links between variations in extraterrestrial energy inputs to the Earth’s atmosphere and meteorological processes has a long, and sometimes controversial, history . Although the total electromagnetic radiation emitted by the sun varies over a range of timescales, with the 11-year cycle being prominent, changes in solar emission at wavelengths that reach the Earth’s lower atmosphere appear to be very small, at least over the period of satellite observations. The variable ultraviolet solar emission is absorbed in the stratosphere and above and leads to small changes in stratospheric temperatures.
In addition to the sun’s electromagnetic radiation, the solar wind contains energetic particle fluxes and magnetic fields of solar origin and generates energetic protons and electrons in the magnetosphere. But these particles are sporadically precipitated, and also deposit their energy at altitudes higher than those where weather phenomena occur. The response of the Earth’s upper atmosphere to these changing energy inputs at the top of the atmosphere includes the optical aurora and the variable electrical current of the auroral electrojets in the polar ionospheres . However, there has been no widely-accepted mechanism to couple known changes in the sun, and the associated changes in the upper atmosphere described above to the region of weather processes within 10 km of the ground.
A less obvious input into the Earth’s atmosphere is the flux of cosmic rays, which are mainly protons of such high energy per particle that they penetrate through the atmosphere to the surface, ionizing nitrogen and oxygen molecules on the way. The flux is modulated by 5-10% by the solar wind magnetic fields, depending on proton energy and latitude, over the 11-year sunspot cycle, and during magnetic storms. Ionization by cosmic rays makes the atmosphere weakly conducting, and allows the global atmospheric electric circuit to form. Thunderstorms and electrified shower clouds, located primarily at low latitudes, generate upward currents and maintain the highly-conducting ionosphere at an elevated electrical potential, typically near 240 kV, relative to the Earth’s surface. This potential gradient drives a downward current density over the entire globe. Horizontal current flow through the solid Earth and oceans back toward low latitudes completes the circuit.
Acting as a catalyst on cloud processes, the atmosphere’s electrical circuit could provide at least one missing link between variations in solar activity and meteorological responses in the troposphere , independent of atmospheric phenomena in the region between about 10 km and the ionosphere. The downward current density appears capable of coupling changing conditions at ionospheric altitudes to meteorological processes near the ground, at least in Polar Regions. Specifically, the electrical conductivities of clouds differ from that of the surrounding clear air, so that when the downward current encounters a cloud, an excess of charges (on air ions, aerosol particles and droplets) of one sign (space charge) develops on the upper and lower cloud boundary regions. Then, the microphysical processes within the cloud can be influenced by the 20% or so variation in the current density and space charge, at least in Polar Regions where the observed cloud characteristics appear to make them susceptible to the presence of such excess charges. These clouds consist of one or more layers of stratus clouds, formed in the electrically clean polar air with its quite low concentrations of aerosols and cloud condensation nuclei (CCN). For these clouds, the optical thickness of the clouds and hence radiative coupling to the ground has been found to vary with the CCN concentration .
The production and loss processes of CCN can be affected by the electrical charges on ions, CCN, and droplets. This leads to slightly altered droplet and ice concentrations , which can then appear as a change in the optical thickness of the clouds, which in turn alters the thermal longwave radiation emitted toward the ground. Through this sequence of events, changes in the ionosphere-to-ground electric current density could be followed, after a time lag characteristic of the processes at work, by a change in downwelling longwave irradiance observed at the ground. Variations in the current density might arise (1) from the thunderstorms and electrified clouds primarily responsible for creating the ionosphere-to-ground potential difference, (2) from changes in the auroral electrojets (horizontal electrical currents flowing at ionospheric altitudes triggered by varying solar energy inputs) which at times pass over the South Pole, and for which effects on current density have been observed in the northern auroral latitudes , or (3) from changes in ionospheric potential induced by a changing solar wind magnetic field moving past the conducting upper atmosphere. The last of these influences is confined to a restricted region, out to about 1000 km from the geomagnetic poles owing to the near vertical geomagnetic field lines there being effectively connected, inside the auroral zones, to the solar wind magnetic field.
The sequence of cause and effect outlined above contains steps which are difficult to quantify. Given this complexity, a useful approach is to carry out statistical studies using atmospheric observations combined with proxies for the ionosphere-to-ground current density. If statistically significant links appear, this increases confidence that the hypothesized physical mechanisms produce effects in the lower atmosphere which are sufficiently large to detect. Downward longwave irradiance observed at the ground is an ideal choice for the atmospheric parameter since it is directly related to the temperature profile and the optical thickness of the overhead clouds that, by their thermal radiation, warm the underlying atmosphere and surface.
Data and Analysis
A long-term dataset of downward-propagating longwave irradiance exists for the South Pole and is available electronically via NOAA’s Earth System Research Laboratory. This work uses 24-hour mean irradiances for the period from late 1993 through mid-2017. The task is to seek a statistical link between a measure of daily longwave irradiance from which the annual cycle has been removed, denoted by y(d), and a proxy variable for the vertical current density, J(d-l). Here d labels a day of the dataset, d=1, 2, …, and l is a time lag in days. Large values of y(d) correspond to the presence of thick clouds whose longwave emission originates from relatively low, warm altitudes. Clear skies, where relatively high, cool altitudes preferentially contribute to thermal emission, lead to smaller observed longwave irradiances.
A suitable regression model is:
y(d) = a0(l) + a1(l)T(d) + a2(l)T(d)2 + a3(l)J(d-l) + ε(d)
where T(d) is a linear trend in time, and inspection of the dataset motivated the quadratic term T(d)2. The residual ε(d) has a mean of 0.0 and includes autocorrelation with a one-day lag. Physically the autocorrelation arises from the persistence of cloudiness. Values for the regression coefficients ai(l), i=0,…,3, are estimated by least-square methods. The coefficient of interest here is a3(l) at time lags between 0 and several days, where the value of l is characteristic of the physical processes at work. The variance in y not explained by the regression model determines a 95% confidence range on the estimate of a3(l).
Various choices for the independent variable J(d-l) allow examining the three different mechanisms that can influence the vertical current density. The first calculation sets J(d-l) equal to daily measurements of the ground-level vertical electric field, adjusted to correct for local meteorological contamination. The data come from the stations Vostok and Concordia on the Antarctic continent . This case addresses the fundamental question of whether polar cloud properties are coupled to the downward electric current density, which, as discussed earlier, is primarily internally-generated by the ionospheric potential due to low-latitude thunderstorms.
Application of the regression model to these datasets produced a statistically significant value of a3(l) at the time lag l=1 day during the dark portion of the polar year. The results imply that two days whose electric fields differ by +22.4 Vm-1 (one standard deviation of the mean field) are followed by days whose longwave irradiances differ by +2.8±1.9% where the error bar defines the 95% confidence range. Here the difference in longwave irradiance has the same sign as that in the electric field. A day with a large electric field is followed the next day by a value of longwave irradiance slightly larger than would exist otherwise. If one interprets this result based on the hypothesized physical mechanism, one concludes that a larger vertical current density, inferred from the larger electric field, tends to increase the opacity of polar clouds over a time scale of one day.
The next case sets J equal to the daily magnetic Ap index which is a measure of variability in the ground-level geomagnetic field. These variations are initiated by changing ionospheric currents such as in the auroral electrojet which also causes changes in the local ionospheric potential and the local downward current density. In this analysis, the regression model shows a statistically significant negative link between longwave irradiance and the Ap index from l=2 days prior. For daylight and dark periods combined, a +10 unit increase in Ap is followed by a longwave irradiance -0.6±0.5% smaller than would occur otherwise. This is consistent with a small decrease in cloud opacity following a period of increased magnetic activity. Note that Ap is a measure of temporal fluctuations in the horizontal electrical currents at ionospheric altitudes. These, in turn, lead to variability in the ionospheric potential and in the ionosphere-to-ground current density imposed on the background due to the other two influences identified above. The change in current density and in the longwave irradiance following an increase in Ap need not have the same sign as the response to either of the other two cases.
The third and last case sets J equal to the daily east-west component of the interplanetary magnetic field obtained from the NASA/Goddard Space Flight Center’s OMNIWeb electronic archive. This regression failed to identify a relationship between longwave irradiance and the magnetic field that met the 95% significance standard. If a coupling of longwave irradiance to the interplanetary magnetic field indeed exists, it is too weak to confirm at the geomagnetic latitude of the South Pole. Note that the geomagnetic pole is displaced by about 10o of latitude from the geographic South Pole. New analyses using longwave radiation acquired closer to the geomagnetic pole, either North or South, are needed to better-define any link to the interplanetary magnetic field. Also note that the correlations described above are with fluctuations in the current density of only 20% or so, and only on the day-to-day timescale. The effect of changes in mean current density, which varies on decadal and longer-term solar cycles with changes in cosmic ray flux, have not been evaluated.
The results of this research are consistent with the hypothesis that the downward atmospheric electrical current influences cloud properties in the Polar Regions and, consequently, longwave irradiance received at the ground. Since the potential difference that drives this current responds to processes occurring at ionospheric altitudes and higher, the atmosphere’s electrical circuit can be responsible for coupling the ionosphere and the near-space environment to meteorological processes near the ground.
Published by John E. Frederick and Brian A. Tinsley
These findings are described in the article entitled The response of longwave radiation at the South Pole to electrical and magnetic variations: Links to meteorological generators and the solar wind, recently published in the Journal of Atmospheric and Solar-Terrestrial Physics (Journal of Atmospheric and Solar-Terrestrial Physics 179 (2018) 214-224). This work was conducted by John E. Frederick from the University of Chicago and Brian A. Tinsley from the University of Texas at Dallas.
- McCormac, B.M., and T.A. Seliga, 1979: Solar-Terrestrial Influences on Weather and Climate, D. Reidel Publishing Co., Dordrecht, Holland, 346 pp.
- Ratcliffe, J.A., 1972: An Introduction to the Ionosphere and Magnetosphere, Cambridge University Press, Cambridge, United Kingdom, 256 pp.
- Tinsley, B.A., 2008: The global atmospheric electric circuit and its effects on cloud microphysics. Rep. Prog. Phys. 71, 066801. https://iopscience.iop.org/article/10.1088/0034-4885/71/6/066801/meta.
- Mauritsen, T., Sedlar, J., and 10 others, 2011: An Arctic CCN-limited cloud-aerosol regime. Atmos. Chem. Phys., 11, 165-173. https://www.atmos-chem-phys.net/11/165/2011/.
- Tinsley, B. A. 2010: Electric charge modulation of aerosol scavenging in clouds: Rate coefficients with Monte Carlo simulation of diffusion, J. Geophys. Res., 115, D23211, doi:10.1029/2010JD014580.
- Frank-Kamenetsky, A. V., Kotikov, A. L., Kruglov, A. A., Burns, G. B., Kleimenova, N. G., Kozyreva, O. V., … & Odzimek, A., 2012: Variations in the near-surface atmospheric electric field at high latitudes and ionospheric potential during geomagnetic perturbations. Geomagn. Aeron. 52(5), 629-638. .
- Burns, G.B., A.V. Frank-Kamenetsky, B.A. Tinsley, W.J.R. French, P. Grigioni, G. Camporeale and E.A. Bering, 2017: Atmospheric global circuit variations from Vostok and Concordia electric field measurements. J. Atmos. Sci. 74, 783–800, http://dx.doi.org/10.1175/JAS_D_16–0159.1.