What if the Earth rotated backwards?

Hi there! I recently realized that over more than 100 articles, I never spoke about our Earth. Of course, you can say that, when I mean planets, I implicitly mean planets other than our Earth… There was probably something like that…
Anyway, our Earth is our home, and as such, it is of the uttermost importance. In particular, the global warming threatens it, and threatens the mankind itself. This is why we must study the Earth, but don’t worry, the Earth is studied.
Today I present simulations of the climate that the Earth would have, if it rotated backwards, at the same rate.
Of course, this is a theoretical study, which does not reproduce a real situation. But this is anyway interesting, because it permits us to understand the role of the different factors, which affect the climate. What is the role of the spin direction?
This is the question this study answers. The study is The climate of a retrograde rotating Earth, by Uwe Mikolajewicz et al., and it has recently been published in Earth System Dynamics.

The climate of our Earth

The climate of our Earth is influences by 4 factors:

  1. the astronomical factors
  2. the atmospheric circulation
  3. the oceanic circulation
  4. the ones I forget

The astronomical factors (axial tilt)

The obliquity of the Earth, or axial tilt, is responsible for the seasons. The rotation axis of our Earth is not orthogonal to its orbital plane around the Sun (the ecliptic), but is tilted by some 23° (somehow the angle between your index and your middle fingers, when you open your hand). The consequence is that the two poles do not see the sunlight six months a year, alternatively. And the other regions of the Earth have varying day durations, which affect the temperature. You have the seasons.

The team of Jacques Laskar (IMCCE, Paris Observatory) has shown that the Moon stabilizes the axial tilt of the Earth (see here). In other words, a moonless Earth would have experienced large variations of the axial tilt, hence large variations of the climate. So large that they may have threatened the development of life on Earth, since we need to adapt to the climate. We can do it when the changes are slow enough… and our fear with global warming is not (only) the warming itself, but its acceleration… Anyway, we are alive thanks to our Moon.

In fact, the astronomic forcing affects the climate on a wider range. The Serbian geophysicist and astronomer Milutin Milanković has hypothesized (and this has been confirmed by several teams since then) that the variations of the orbit and the rotation of the Earth were responsible for the paleoclimates. This theory is now known as the Milanković cycles.

But astronomic forcing is not everything. This affects the insolation of a given place, providing some energy to heat the Earth (not the whole energy actually, but let us neglect this point). Once a planet is illuminated, it responds… and the response depends on its constituents, the atmosphere playing a critical role.

The atmospheric circulation

As you know, our Earth is surrounded by an atmosphere, which is a layer of air, mostly composed of nitrogen and oxygen. Its pressure decreases with the altitude, 3 quarters of it being in the 11 lowest kilometers, while the boundary at the atmosphere is considered to be at about 100 km. This atmosphere is responsible for greenhouse effect, which heats the surface. It also increases the pressure, this permits the existence of liquid water. Moreover, it protects us from ultraviolet radiation, meteorites (many of them being fragmented when encountering the atmosphere), and allows us to breath. You can forget life on an atmosphereless Earth.

Beside this, the atmospheric circulation redistributes the thermal energy on Earth. You know the winds.
More precisely, this circulation is structured as cells, which take hot air at given locations of the surface, before releasing it back somewhere else. The main effect is due to latitudinal cells (Hadley, Ferrel, and polar cells), which permit heat transfers between different latitudes, but there is also a longitudinal motion, known as zonal overturning circulation.

Oceans play a key role in the regulation of our climate, since they have a kind of thermal inertia, which affects the temperature of the coastal areas.

The oceanic circulation

I mean the oceanic currents, which are water displacements. This may transfer hot water to colder regions, and conversely. An example is the North Atlantic Drift, aka Gulf Stream, which is responsible for the pretty moderate winters in Europe, while Canada freezes. There are also currents designated as gyres, since they have a pretty circular motion on a very large scale.
Moreover, you also have formation of water masses in the Atlantic, i.e. masses of water, which properties (temperature, salinity,…) are pretty homogeneous, and different from the surrounding waters.

Atmospheric and oceanic circulations are influence by the Coriolis effect, which is the consequence of the Earth rotation… and this study is on the influence of the Earth rotation.

The ones I forget

Sorry, I don’t remember 🙂

Let me mention anyway the influence of the land, which of course blocks the oceanic currents, and also may affect the atmospheric ones, in particular if you have mountains.

Different climates

All of these effects make meteorology a very complicated science. And you also have different climates on Earth, such as (following Köppen climate classification):

  • tropical climates (constant high temperatures),
  • dry climates (deserts),
  • temperate climates,
  • continental climates, where you have large variations of temperature between summer and winter,
  • and polar climates (the coldest ones).

You cannot pretend simulating the climate of the Earth if you don’t get these 5 climates.

The Max Planck Institute Earth System Model

The authors are experts in climate simulation. This is a very difficult task, since you have to implement the interactions between all the physical parameters (insolation, oceanic currents, atmospheric circulation,…), in a code which is non-linear and depends on multiple variables. Basically, when an equation is non-linear, you cannot simply derive its solution. Instead, you need to integrate the equation numerically, and the solution may be very sensitive to your parameters, your initial conditions (how is the climate when you start the simulations?), and your numerical scheme.

In particular, you split the atmosphere and the oceans on a grid of finite elements, and your numerical code simulates the solution element by element, time after time. This requires high performance computing tools.

The authors dispose of a dedicated numerical model, the Max Planck Institute Earth System Model (MPI-ESM), which couples the atmosphere, ocean and land surface through the exchange of energy, momentum, water and carbon dioxide. This homemade tool has been developed after years of study. It interfaces the simulations of different physical processes, all of them having been developed and improved since many years.
The authors have used the MPI-ESM many times in the past, which makes it reliable.

Intensive numerical simulations

In present study, the authors ran two sets of simulations:

  • CNTRL, which are consistent with our knowledge of the Earth,
  • and RETRO. To each CNTRL simulation corresponds a RETRO one, in which the Earth rotates backwards.

Each set is composed of 1,850 climate conditions (i.e. 1,850 different simulations), over 6,990 years. The authors point out that the simulations should be over a long enough duration, to permit the climate to reach an equilibrium state. The simulations show that in practice, the equilibrium is reached in some 2,000 years.

CNTRL simulations are necessary since, if you just compare a RETRO simulation with our observed climate, you cannot be sure whether the difference comes from the retrograde rotation, or from an effect which would have been inaccurately modeled. Moreover, running so many simulations permits to distinguish robust solutions, which give in some sense the same climate for many simulations, from anecdotic ones, i.e. due to particular initial conditions. Such a non-linear system of equations (Navier-Stokes, etc.) may be chaotic, which implies to be possibly very sensitive to the initial conditions, in a given range which we do not really know…

In the RETRO simulations, the backward rotation is modeled as:

  • the inversion of the Coriolis parameter in the oceanic and atmospheric circulations,
  • the inversion of the Sun’s diurnal march in the calculations of radiative transfer.

And one the simulations have run, they get the results. The question you may ask is: would that affect the global temperature of the Earth? It appears that no. You have no change on average, I mean the mean temperature remains pretty the same, but you have dramatic local changes. Let me emphasize two of them.

The Atlantic and the Pacific exchange their roles

As you can imagine, the inversion of the rotation results in inversion of the oceanic currents and the zonal winds. No need to run the simulations to predict this. But the simulations show unexpected things.

The Atlantic ocean is known for its water masses, and the CNTRL simulations get them. However, the RETRO simulations do not have them in the Atlantic, but in the Pacific Ocean.

A green Sahara

Another change is that the monsoons occur in the Sahara and Arabian Peninsula. This dry area, made of desert, would be a forest if the Earth rotated backwards! However, the world’s biggest desert would have been in the Southern Brazil and Argentina.

You can finally ask: why the authors did this study, since a backward rotating Earth is not realistic? Just because we need to fully understand the climate, and the rotation direction is one of the effects affecting it. We do not know whether this could apply to an extrasolar planet, or whether the results would help us to understand something else… That’s research, but trust me, it is useful one! Climate science has become a critical topic.

The study and its authors

And now, the authors:

And that’s it for today! Please do not forget to comment. You can also subscribe to the RSS feed, and follow me on Twitter, Facebook, Instagram, and Pinterest.

Big impact on Ceres

Hi there! Today: Ceres. (1) Ceres is the largest object of the main asteroid belt, so large that the International Astronomical Union (IAU) classified it as a dwarf planet in 2006. As many planetary bodies, it is craterized, the largest crater being named Kerwan. This crater has a diameter of 280 km. But this is not the most remarkable one. The crater Occator, which diameter almost reaches 100 km, is particularly interesting since it exhibits bright spots, which are probably the signature of past hydrothermal activity. This raises the interest of the scientific community, since it could reveal a geophysical activity and water below the surface.
The study I present, The various ages of Occator crater, Ceres: Results of a comprehensive synthesis approach, by A. Neesemann et al., tries to be as accurate as possible on the age of Occator, in summarizing the previous studies and in using as many data as possible. These are actually data provided by the spacecraft Dawn. This paper will be published in Icarus soon.

The dwarf planet (1)Ceres


The quest for an object between the orbits of Mars and Jupiter was initially motivated by the Titius-Bode law. This empirical law, which is now proven to be absolutely wrong, noticed a arithmetic progression between the orbital radii of the known planets, and was confirmed by the discovery of Uranus in 1781 (however, it is inconsistent with the presence of Neptune). Anyway, this convinced former astronomers that something was there, and it revealed to be true. A group led by Franz Xaver von Zach looked for an object with a semimajor axis close to 2.8 AU (astronomical units, remember that 1 AU is 150 million kilometers, which is the orbital radius of our Earth). But that group did not discover Ceres.

Ceres has been serendipitously discovered in 1801 by the Italian astronomer Giuseppe Piazzi in Palermo, Sicilia. He wanted to observe the star HR 1110, but saw a slowly moving object instead. He noticed that it looked somehow like a comet, but that it was probably better than that. Ceres was found!

Giuseppe Piazzi (1746-1826) pointing at Ceres. © Palermo Observatory
Giuseppe Piazzi (1746-1826) pointing at Ceres. © Palermo Observatory

Later, the group led by von Zach discovered many asteroids. One of them, Heinrich Olbers, is credited for the discoveries of Pallas, Vesta, and the periodic comet 13P/Olbers. He also gave his name to the Olbers paradox, which wonders why the night is so dark while we are surrounded by so many stars.


You can find below some of the orbital and physical properties of Ceres.

Semimajor axis 2.77 AU
Eccentricity 0.075
Inclination 10.6°
Revolution 4.60 yr
Rotation 9 h 4 min
Diameters (965.2 × 961.2 × 891.2) km
Density 2.161 g/cm3

These orbital elements and its size make it the largest object of the main asteroid belt. You can see a small eccentricity, and a pretty fast rotation period with respect to its orbital one (i.e., the revolution). Moreover, its equatorial section is pretty circular, i.e. if you look at its 3 diameters, the two largest ones of them are very close, and in fact the uncertainties on the measurements are even consistent with a strict equality. However, the polar diameter is much smaller. This is a consequence of its rotation, which flattens the body.

You can also notice a density, which is between the one of the water (1) and the one of silicates (3.3). This means that its composition should be a mixture of both, i.e. silicates and water ice.

Ceres seen by Dawn © NASA/JPL-Caltech/UCLA/MPS/DLR/IDA
Ceres seen by Dawn © NASA/JPL-Caltech/UCLA/MPS/DLR/IDA

The physical properties and the image above are due to the spacecraft Dawn. This mission is close to its end.

Dawn at Ceres

The spacecraft Dawn has been launched from Cape Canaveral, Florida, in September 2007, and reached the asteroid Vesta in 2011. After a little more than one year in orbit around Vesta, it left it and has been trapped by the gravity field of Ceres in March 2015. This mission will be completed soon.

Dawn consists of three instruments:

  • the Gamma Ray and Neutron Detector (GRaND) Instrument,
  • the Visible and Infrared Spectrometer (VIR) Instrument,
  • and the Framing Camera (FC).

Dawn is essentially an American mission, even if Germany provided the Framing Camera. The German study we discuss today uses FC data.

The orbital journey of Dawn around Ceres consists of several phases, which are different orbits. This results in variable resolutions of the images. The prime mission considered two mapping orbits, the HAMO (High Altitude Mapping Orbit) and the LAMO (for Low Altitude), at distances of 1,470 and 375 km of the surface, respectively. Since then, the mission has been extended, and the spacecraft is now at only 50 kilometers of the surface. High resolution expected.

This mapping orbits permitted to map comprehensively the surface of Ceres. Unsurprisingly, that survey revealed many craters.
We are today interested in Occator, which is not the largest one, but contains bright spots, possibly signatures of a recent hydrothermal activity.

Occator crater

Occator crater is located in the northern hemisphere of Ceres. Its diameter is some 90 km, which does not make it the largest one, but it is particularly interesting for the bright spots it shows. To be honest, there are bright spots at other locations of Ceres, but anyway Occator is remarkable for that. The spot in the center is a dome called Cerealia Facula, while the small spots are called the Vinalia Faculae. You can see them below, on these high-resolution images due to the extended mission.

Occator Crater on Ceres, with its central bright area called Cerealia Facula. © NASA/JPL-Caltech/UCLA/MPS/DLR/IDA/PSI
Occator Crater on Ceres, with its central bright area called Cerealia Facula. © NASA/JPL-Caltech/UCLA/MPS/DLR/IDA/PSI
Vinalia Faculae © NASA/JPL-Caltech/UCLA/MPS/DLR/IDA/PSI
Vinalia Faculae © NASA/JPL-Caltech/UCLA/MPS/DLR/IDA/PSI

Yes, there are domes, due to hydrothermal activity! You can find interesting analogies with Earth features here. But basically, the two possible explanations are for now:

  1. either the heat from the impact that formed the crater caused briny liquid or mushy ice to push up on the surface,
  2. or the heat from the impact could have enhanced activity related to pre-existing liquid reservoirs just below the surface.

Anyway, this reveals water! And this makes Ceres and its crater Occator fascinating.

How a crater evolves

This study wants to estimate the age of Occator, or in other words, date the impact that created it. For that, you examine its current state, and guess how long it took from the impact to the observed state.

Because of the elasticity of the surface, after some time (hundreds of millions years, maybe a little more…) the surface relaxes. The consequence is that the crater gets less deep, and its slopes get gentler. A 3-D terrain model will give you the numbers. But the dynamics of the relaxation process is barely constrained.

Another evolution is that the crater is covered by something else. This something could be other, more recent craters. If the new crater is larger than the older one, then the oldest disappears. However, if the new crater is small with respect to the old one, then you see both, and by counting the small craters, you can say “it took this time to get so many craters, so the age is at least…”. OK

But how to constrain this? You calibrate your models with better known bodies, like the Moon, and / or dynamical models of the bombardments. Previous studies have used Lunar Derived Models and Asteroid-flux Models… of course with different outcomes.

In the specific case of Occator, the hydrothermal activity revealed by the bright spots has generated ejecta blankets, as lobate deposits.

Previous estimations of Occator’s age

The quest for the age of Occator crater began with the first data on Ceres, i.e. in 2015. Here are the already published numbers

  • Nathues et al. 2015: 78 ± 5 Ma (million years). This measurement is based on crater counting, and only HAMO data. In particular, the more accurate low-altitude data were missing at that time,
  • Nathues et al. 2016: 6.9 Ma, based on the interior lobate deposits,
  • Jaumann et al. 2016: between 100 and 200 Ma, depending on how you calibrate the dating from craters,
  • Nathues et al. 2017: 34 ± 2 Ma, from the creation of the central dome, i.e. Cerealia Facula,
  • Nathues et al. 2018, stated that the dispersed bright deposits Vinalia Faculae were younger than 2 Ma, in using low-altitude high-resolution images.

The study we now discuss uses almost all of the data, and so should be more accurate.

A young crater anyway

It is interesting that a study points out all of the possible numbers, given the models, the data, and the physical process considered (crater counting, age of ejecta,…). In particular, if the hydrothermal activity has been triggered by the impact which created Occator, then dating the ejecta should tell us something accurate.

The authors find an age of 21.9 ± 0.7 Ma for the crater in using the Lunar Derived Model, and between 1 and 64 Ma in using the Asteroid-flux Derived Model. You see, lots of uncertainties… as they say, the model ages are a matter of perspective. But anyway, this is a very young and interesting crater!

The study and its authors

And that’s it for today! Please do not forget to comment. You can also subscribe to the RSS feed, and follow me on Twitter, Facebook, Instagram, and Pinterest.

A blue comet

Hi there!

Today we discuss about the comet C/2016 R2. This comet has been discovered in September 2016, i.e. some 20 months before its closest approach to the Sun (passage at perihelion). Meanwhile, it heated more and more, and outgassed. This permitted many observations of the comet and its tails, and it appears that this comet is different from the classical ones. In what way? You have to read until the end.
I present The extraordinary composition of the blue comet C/2016 R2 (PanSTARRS), by N.Biver, D.Bockelée-Morvan, G.Paubert, R.Moreno, J.Crovisier, J.Boissier, E.Bertrand, H.Boussier, F.Kugel, A.McKay, N.Dello Russo and M.A.DiSanti. This study has recently been accepted for publication in Astronomy and Astrophysics.

The comet C/2016 R2

This comet has been discovered on September 7, 2016 at 6.3 astronomical units from the Sun, on the telescope Pan-STARRS (Panoramic Survey Telescope And Rapid Response System), which is located on the Haleakalā (Maui, Hawaii, USA). It was then moving toward its perihelion, i.e. toward the Sun. The passage occurred in May 2018, at a distance of 2.06 AU from the Sun. Since then, it is moving away, and it should come back in about 20,000 years… if it comes back. You can find below its orbital elements, taken from the JPL Small-Body Database Browser. These elements are computed from almost 3,000 observations, over 2 years (remember: the comet has been discovered in September 2016).

Semimajor axis 736.30 AU
Eccentricity 0.996
Inclination 58.22°
Perihelion distance 2.60 AU
Period 19,979.85 yr

This is what we call a long-period comets. You have comets of the Jupiter family, with period of a few years, you have comets with periods close to the century, like the very famous Halley comet, which period is 75 years (next passage in July 2061, be ready), and you have long-period comets like this one.

In fact, we have no proof that this comet already visited us 20,000 years ago, and we cannot be sure it will in 20,000 years. Look at the eccentricity: 0.996 is huge for an elliptic orbit, it actually means that the orbit is almost parabolic. In other words, it is close to never come back. Planetary perturbations or even a star passing by could be strong enough to destabilize the orbit of C/2016 R2, and make it leave our Solar System for ever. Conversely, it could make it more stable, and we can even imagine that this comet has an extrasolar origin. It could have been formed around another star, having been ejected from that system, having visited us, and been stabilized around the Sun. Which does not preclude a future ejection.
In fact, it is difficult to know, since such elongated orbits and such small objects are very sensitive to small planetary perturbations.

You can get clues on the origin of a planetary body by studying its composition. This may be pretty easy (or I should say: not that difficult) because

  1. amateurs are usually enthusiastic with comets,
  2. when a comet approaches the Sun, it creates beautiful tails.

When you observe the tails, you have part of the composition.

Spectroscopy of a comet

As discovered Fred Whipple, a comet is a dirty snowball, which is composed of a nucleus, and tails. This has been confirmed in 1986, when the comet 1P/Halley visited us. The nucleus is composed of water ice and silicates.

Approaching the Sun heats the comet, and sublimates its components. This is how it creates 3 tails:

  1. a dusty tail, which is the visible one. It is curved, and located behind the comet,
  2. an ion tail, which is made of gases. Its direction is opposite to the one of the Sun, because it is strongly affected by the Solar wind,
  3. a weak dusty antitail, which points opposite to the dust tail.

And for guessing the composition of these tails (you can also say the coma), you use spectroscopy. Basically:

  1. the Sun lights the tail,
  2. the Solar light is made of a wide spectrum of radiations. The radiations, which are characteristics of the elements present in the coma, are blocked.
  3. you observe the spectrum of the light crossing the coma. From this spectrum, you know which elements are present.

Of course, this is a little more complicated than that. First, you have to consider that the light you observe crosses the atmosphere, which affects it. So, you have to remove this effect. And then, you also have to consider that a cometary activity might be a weak process (depending on the perihelion distance and on the size of the nucleus), and some elements are sometimes observed, sometimes not. Fortunately, several teams have observed the comet, which secures the results.

For observing these spectra, you need dedicated facilities, which do not necessary observe the visible light. Let us present them now.

Many observation facilities

I here restrict to facilities used by the authors of that specific study, but there are many more.


IRAM is a French acronym for Institut de RAdioastronomie Millimétrique, or German for Institut für Radioastronomie im Millimeterbereich, or even Spanish for Instituto de Radioastronomía Milimétrica, i.e. institute of millimetric radioastronomy. What does that mean? It means that this facility observes in wavelengths, which are close to the millimeter. Remember that the visible light lies between 430 (for the red) and 700 (for the violet) nanometers, or 0.0007 millimeters. So, the human eye is definitely not sensitive to millimetric emissions. This part of the spectrum approaches the one of the radio waves.

IRAM is based in Grenoble, France, but Germany and Spain are also involved. Beside this, IRAM has worldwide collaborations.

It has 2 observing facilities:

  1. a 30 m radio telescope located on Pico Veleta (Andalucia, Spain),
  2. an interferometer, located on the Plateau de Bure, in the French Alps.

Both sites are at high altitude, i.e. 2,850 m for the Spanish site, and 2,550 m for the French one. IRAM has been created in 1979, and is still upgrading its facilities. Spain joined it in 1990, and the Plateau de Bure Interferometer, which was originally composed of 6 antennae, has now 10 and should have 12 by 2020, in the framework of the project NOEMA, for NOrthern Extended Millimeter Array. Its deployment, i.e. the inauguration of the seventh antenna, started in September 2014.

IRAM's 30m telescope at Pico Veleta © IRAM
IRAM’s 30m telescope at Pico Veleta © IRAM

The authors used the 30-m telescope based in Spain, to target the carbon monoxide, the hydrogen cyanide HCN, and the hydrogen sulfide H2S, at the frequencies 231, 266 and 169 GHz, respectively. The observations were conducted during two evenings, on 23 and 24 January 2018.


They also used the Nançay radio telescope. This facility is based in the center of France, and depends partly on Paris Observatory. The radio telescope observes centimetric waves. For that, it is composed of 2 mirrors, one is planar while the other one is spherical. These are mirrors for centimetric waves, i.e. these are actually railings, in which the metal bars are spaced by less than 1 centimeter. It mainly observes the wavelengths 21, 18 and 9 cm, which correspond to hydrogen, hydroxide OH, and the methylidyne radical CH.

The radio telescope, by <a href="//commons.wikimedia.org/wiki/User:Julien_Descloux" title="User:Julien Descloux">Julien Descloux</a> — <span class="int-own-work" lang="fr">Travail personnel</span>, <a href="https://creativecommons.org/licenses/by-sa/3.0" title="Creative Commons Attribution-Share Alike 3.0">CC BY-SA 3.0</a>, <a href="https://commons.wikimedia.org/w/index.php?curid=8339640">URL</a>
The radio telescope, by Julien DesclouxTravail personnel, CC BY-SA 3.0, URL

Beside this, Nançay has also a decametric interferometer composed of 144 antennae, and an antenna belonging to the Low-Frequency Array (LOFAR) network, which observes metric wavelengths.

The authors used the radio telescope between January and March 2018, for about one hour on average every 2 days.

Amateurs observations

These observations were supplemented by optical observations conducted by amateurs, who co-author the study.
The analysis of visible light could permit to detect, for instance, carbon monoxide CO (to confirm millimetric observations), the cyano radical CN, or the dinitrogen N2.

Detected species

First, the overall emission of the comet constrains the temperature of the gas. For C/2016 R2, it should be close to 23 K (-250°C, or -418°F).

Regarding the species: the authors detected a very large production of carbon monoxide CO. This, combined with a very low dust production (with respect to known comets), makes the coma to be blue, instead of tending to be red or yellow. And relatively to CO, there is a strong depletion of water H20, methanol CH3OH, formaldehyde H2CO, hydrogen cyanide HCN, and hydrogen sulfide H2S.

However, and this is very surprising, the authors detected an excess of dinitrogen N2, with a ratio N2/CO close to 0.08. In this case, N2 dominates the nitrogen budget.

Why so much dinitrogen

We don’t know! And this is why it is interesting.

Such an abundance of dinitrogen is very unusual. The only previous detection of dinitrogen was in situ, by Rosetta on 67P/Churyumov-Gerasimenko. In that case the ratio N2/CO was about 0.006, i.e. 10 times lower than for C/2016 R2. And for the other comets: just no detection.

This means that this comet did not form with the other comets. It formed elsewhere. And this makes this comet unique.

The study and its authors

And that’s it for today! Please do not forget to comment. You can also subscribe to the RSS feed, and follow me on Twitter, Facebook, Instagram, and Pinterest.