I assess planetary habitability by developing atmospheric models of terrestrial planets both inside and outside the solar system.  I use my simulations to inform astronomers which exoplanets are most likely to be habitable. I also study the habitability of the terrestrial planets in our own solar system (including Mars, Venus, and our own Earth) and use this knowledge to improve our understanding of exoplanet habitability. My recent work has focused on the evolution of the habitable zone, the habitability of ocean worlds, atmospheric escape/erosion processes, and early Mars paleoclimate. The subsections within describe my research interests in greater detail.


Habitable Zones of Main-Sequence Stars

The habitable zone (HZ) is defined as the region around the star where water can be sustained as a liquid on a planetary surface (Kasting et al., 1993). The conservative boundaries of the HZ are defined by the moist greenhouse limit at the inner edge and the maximum greenhouse limit on the outer edge (ibid). I have developed an improved 1-D radiative-convective climate model that was used to recalculate the HZ boundaries for F-M main sequence stars (Kopparapu and Ramirez et al., 2013). For our solar system, the HZ boundaries are at 0.99 – 1.67 AU, as compared to the 0.95 – 1.67 AU obtained by Kasting et al. (1993).


The classical CO2-H2O habitable zone (blue) with volcanic hydrogen extension (red) for main-sequence stars with effective temperatures ranging from 2,600 – 10, 000 K.

I then applied the CO2-H2 greenhouse from my early Mars work to demonstrate that volcanic outgassing of atmospheric hydrogen can extend the outer edge of the habitable zone in our solar system from 1.67 AuU to 2.4 AU (Ramirez and Kaltengger, 2017). The width of the habitable zone is similarly extended for all spectral classes (A – M).  Unlike primordial hydrogen accreted from the protoplanetary disk, as suggested in a previous study, which is  gone within a few million to tens of millions of years, so long as volcanic outgassing of hydrogen outpaces its escape to space, such planets can remain habitable for geologically-significant timescales (~500 million years – 1 billion years or more). Hydrogen is a very light gas, though, and tends to escape to space very quickly. However, worlds can remain habitable in this new volcanic hydrogen habitable zone so long as hydrogen outgassed from volcanoes outpaces its escape to space.

The extra heat from hydrogen also means that considerably less CO2 is needed to support habitable conditions.  Moreover, the more hydrogen is supported in such atmospheres, the larger the atmospheric scale height, facilitating the detection of potential bioindicators by upcoming missions.

Another gas that can be used to extend the habitable zone outward is methane. However, methane is unusual in that it can be an effective greenhouse gas on planets orbiting hotter stars (> ~ 4500 K). whereas it is an anti-greenhouse gas on worlds orbiting colder stars. Adding ~10% CH4 to an outer edge planet’s atmosphere can increase the  outer edge distance by ~20% if orbiting a hot stars, whereas this amount of methane can decrease the outer edge distance by a similar percentage around colder stars (Ramirez and Kaltenegger, 2018).

We also find that dense CO2-CH4 atmospheres near the outer edge of hotter stars potentially signal inhabitance. Therefore, in Ramirez and Kaltenegger (2018), we argue that such worlds should be explored if found.


The classical CO2-H2O habitable zone (blue) with methane extension (light brown) for main-sequence stars with effective temperatures ranging from 2,600 – 10, 000 K.

Habitable Zones of Pre-Main-Sequence Stars

In Kopparapu and Ramirez et al., 2013, the boundaries assumed for all stars are that for the present (4.5 Gyr)  HZ.  However, a star’s luminosity changes greatly over time, especially during the pre-main-sequence, which means that the HZ is also temporally varying. This is important because applying the present day (4.5 Gyr) habitable zone to stars of various ages is almost certainly wrong. Ramirez and Kaltenegger (2014) calculate the pre-main-sequence HZ boundaries as a function of distance and time for F – M stars.

Moreover, the spatial distribution of liquid water and its change during the pre-main-sequence phase of protoplanetary systems is important to understanding how planets become habitable. Such worlds are interesting targets for future missions because the coolest stars could provide habitable conditions for up to 2.5 billion years post-accretion. Moreover, for a given star type, planetary systems are more easily resolved because of higher pre-main-sequence stellar luminosities, resulting in larger planet-star separation for cool stars than is the case for the traditional main-sequence (MS) habitable zone (HZ). We use our 1-D radiative-convective climate and stellar evolutionary models to calculate pre-main-sequence HZ distances for F1-M8 stellar types. We also show that accreting planets that are later located in the traditional MS HZ orbiting stars cooler than a K5 (including the full range of M-stars) receive stellar fluxes that exceed the runaway greenhouse threshold, and thus may lose substantial amounts of water initially delivered to them. We predict that M-star planets need to initially accrete more water than Earth did or, alternatively, have additional water delivered later during the long pre-MS phase to remain habitable. Our findings are also consistent with recent claims that Venus lost its water during accretion.


The pre-main-sequence habitable zone is much wider than the main-sequence habitable zone,  which facilitates planet detection in the former.

                                       Habitable Zones of Post-Main-Sequence Stars

Whereas Ramirez and Kaltenegger (2014) assessed the evolution of the habitable zone during the pre-main-sequence evolution, Ramirez and Kaltenegger (2016) continue the temporal evolution of the habitable zone with the post-main-sequence. As with the previous study, stellar luminosity changes greatly during the post-main-sequence, but in the opposite direction, with stars getting bigger and brighter with time. A star that grows large enough to enter this post-main-sequence is called a red giant.

In this study, we extended the stellar temperature limit of our  radiative-convective climate model from 7,200 to up 10,00 K, allowing habitable zones to be calculated for A – M spectral classes. This updated climate model was used to compute red giant habitable zones for A5 – M1 star types. However, as with the previous work, there are major challenges to habitability that planets have to face. During this time the star is rapidly losing mass, resulting in large stellar winds that erode planetary atmospheres throughout the entire stellar system. As the stellar gravity decreases, planetary orbits will move outward in order to conserve angular momentum.  Once this red giant stage ends in our solar system, Venus and Mercury (possibly Earth as well) would have been engulfed by the Sun, whereas what’s left of Mars’ tenuous atmosphere  would have been completely removed.

In spite of this apparent doom and gloom, we show that planets that are far enough away from their stars are able to at least partially, if not mostly, retain their atmospheres. Our work shows that planets are more  likely to retain their atmospheres around the more massive stars. Furthermore, planets in the habitable zones of the smallest stars (K5 and M1) can stay there up to a few billion years, making them great potential targets for future missions. Finally, our study raises the possibility that Europa analogues in other stellar systems may melt once they enter the post-main-sequence habitable zone, potentially unveiling  any pre-existing life within a subsurface ocean.


The habitable zone moves outward as the red giant Sun grows larger and larger with time. Planetary orbits move outward as the Sun loses mass.


Can Ocean Worlds be Habitable?

It is thought that the carbonate-silicate cycle that planets need to maintain their habitability requires that that the planets contain some land in order to operate. This suggests that planets that are completely engulfed in water (i.e. ocean worlds) may not be habitable (Abbot et al. 2012). Moreover, the resultant high-pressure sea ice that is thought to develop on ocean worlds with inventories approximately 100 (or more) times that of Earth’s is thought to inhibit such a cycle or any sort of sustained volcanism from occurring.

However, Levi et al. (2017) have argued that such ocean worlds would exhibit a new stabilizing feedback cycle in which outgassing of carbon dioxide (with the source originating within the deep ocean floor) is balanced by a net ocean influx of atmospheric carbon dioxide.  Life would be able to to originate on the sea ice within the clathrates that would readily form within the resultant high pressure atmospheres (ibid).

In Ramirez and Levi (2018) we used our coupled single-column radiative-convective and latitudinally-dependent energy balance climate models and determined that the geophysical cycle developed in Levi et al. (2017) may be plausible, so long as planetary day lengths are ~8 hours or less. Such rapid rotation rates are necessary in order to produce the cold icy polar-subpolar and warm tropical-sub-tropical regions needed for such a cycle to operate. We then compute a circumstellar region – akin to the habitable zone- in which such potentially habitable ocean worlds may exist.


Schematic of the Levi et al. (2017) atmospheric-ocean circulation that maintains habitability on ocean worlds (Image Credit: Amit Levi from our work in Ramirez and Levi, 2018)


Early Mars atmospheric evolution

The presence of valley networks on the surface of Mars strongly suggests that the Red Planet was warm enough to exhibit liquid water on its surface at or before 3.8 billion years ago (i.e. Craddock and Howard, 2002). However, climate models have had difficulty producing a feasible greenhouse climate scenario which would have kept early Mars warm with the fainter young Sun. We propose that a CO2-H2 greenhouse could have done the trick (Ramirez et al., 2014a). I am also interested in demonstrating that leading cold Mars hypotheses such as the impact or the icy highland hypotheses (i.e. Segura et al., 2008, Wordsworth et al., 2013) cannot explain the observed surface geomorphology.

We have written a review summarizing the geological and climatological evidence (Ramirez and Craddock, 2018), concluding that it does not support icy scenarios for early Mars, nor transient warming episodes in icy climates. The observations suggest that a warm and semi-arid climate was required to explain valley network formation.


Artist’s impression of a wet and warm early Mars

                                                          Cirrus Clouds to Warm early Mars

Using clouds to warm early Mars is another idea that is considered in warm early Mars simulations. Urata and Toon (2013) have argued that cirrus cloud decks composed of 10-micron (or larger)  sized cloud particles could have warmed the Red Planet. We (Ramirez and Kasting, 2017) had originally shown that cirrus cloud decks could indeed have warmed early Mars but only at extremely high cloud fractions exceeding 75% and if cloud parameters are chosen optimally. I later showed that cirrus cloud fractions as low as 50%, with the addition of hydrogen, can provide warm conditions for early Mars although such cloud fractions are still unrealistically high (Ramirez, 2017).

                                      The Challenges with Transient Warming Models

I have shown (Ramirez, 2017) that the reflectivity of surface greatly increases the difficulty to transiently warm an already glaciated early Martian surface, requiring surface pressures that are ~10 – 60% higher than in the ice-free case. At higher pressures than these, atmospheres are prone to collapse altogether.  Also, according to recent mineralogical studies, surface temperatures may need to well exceed 300 K (298 – 323 K) for the observed clay distribution to be formed under short transient episodes. However, such high temperatures would require *very* high atmospheric pressures (exceeding 5 or even 10 bar) in most cases, which exceed available paleopressure constraints (<= 1.8 bar; Kite et al., 2014; Hu et al., 2015). The geology is clear that terrains on early Mars were not glaciated, which suggests that the climate must have been warm (possibly semi-arid) to explain the surface geology that we observe.


Runaway Greenhouses

The value in studying Earth to learn more about other planets cannot be overstated. In Kopparapu and Ramirez et al. (2013), we calculate the inner edge to be precariously close to Earth’s orbit (0.99 AU). However, we had assumed fully-saturated atmospheres, which overestimates water vapor absorption for Earth. In our most recent work (Ramirez et al., 2014b), we utilize a more realistic relative humidity distribution to address whether or not Earth is truly susceptible to either a moist or runaway greenhouse.


The Earth

Early Earth

study by NASA researchers suggests that solar storms on a very young active Sun ~4 billion  years ago could have initiated interesting nitrogen- based atmospheric chemistry on the early Earth, leading to the production of the greenhouse gas, N2O, and HCN, a vital compound for the origin of life. I summarize their work in my  News and Views and offer some interesting perspectives.  In particular, future work should address what this means for 1) early Mars paleoclimate and 2) assess how effective stellar storms are in triggering nitrogen-based atmospheric chemistry on planets around other stars. Finally, atmospheric scientists and biologists should work  together to constrain how much HCN is needed to fertilize surface biology.


Artist’s impression of frequent and powerful solar storms that would have impacted Earth’s magnetic field nearly 4 billion years ago. Image Credit: Tulsi Voralia