2014 Annual Science Report

VPL at University of Washington Reporting  |  SEP 2013 – DEC 2014

Understanding Past Earth Environments

Project Summary

This year, this interdisciplinary effort continued on two major fronts. First, we furthered the development and use of new techniques that help us characterize environmental conditions on ancient Earth. This included progress on our development of a technique for estimating the atmospheric pressure on Archean Earth, and the development and use other techniques for analyzing the chemistry of Archean lakes. We also used our existing models of ancient Earth to simulate other conditions consistent with the conclusions reached from these laboratory analyses.

4 Institutions
3 Teams
6 Publications
1 Field Site
Field Sites

Project Progress

This year, this interdisciplinary effort continued on two major fronts. First, we furthered the development and use of proxies for the environmental conditions on ancient Earth. This included progress on techniques to constrain the atmospheric pressure on Archean Earth, and the development and use of isotopic proxies for the chemistry of Archean lakes. The second front encompassed modeling of the environments of ancient Earth, with simulations that were constrained by our measurements of the environmental conditions.

Two critical and unconstrained parameters for Archean Earth’s environment are the atmospheric pressure and the alkalinity of lakes. We continued to develop an independent proxy for pressure to complement our previous work (Som et al., 2012). This prior limit on Archean surface atmospheric pressure has led to new climate models that incorporate this constraint on atmospheric pressure (see below).

We also conducted N, S and Se isotopic analyses to investigate the oxygenation state of early Earth environments. The most notable outcome was the discovery, based on extreme nitrogen isotope enrichments, of alkaline lakes in the Archean (Stüeken et al., 2015). The N isotope enrichment was caused by ammonia being released to the atmosphere at pH>9 and, when compared to nitrogen isotope systematics of contemporary marine sediments, shows that Archean oceans were not alkaline, contrary to the “early soda ocean” hypothesis. This also will have implications for our future models, as the pH of these waters will affect the free energy for certain chemosynthetic energy pathways. The effects of this will be included in the “energy-based ecosystem models” currently being developed as part of Task D.

Our modeling efforts leveraged these – and other proxies – to constrain the nature of Archean environments. A significant advance in our ability to do this came from the work led by Mark Claire, who increased the wavelength resolution of our photochemical models, allowing more direct simulation of mass-independent sulfur isotope fractionations (Claire et al., 2014; Figure 1). This has allowed us to re-assess hypotheses on the formation mechanisms for this signal in the rock record, with the potential to further constrain Archean atmospheric composition with laboratory data we expect from other groups this year.

Incorporating these proxies also affected our simulations of Archean climate. We tested atmospheres with CO2, CH4, and H2 using a pressure level varying from 0.25 to 2 bars. We determined that H2 forcings were essentially negligible at the lower atmospheric pressures, because H2 warming comes from collision-induced absorption and at low pressures there are fewer collisions, resulting in a smaller greenhouse effect. However, we found that at these lower pressures, it is possible to keep the atmosphere warm with at least 3% CO2, and 0.15% CH4. These results are warmer than previous models under similar conditions, as a result of updated k-coefficients. These new k-coefficients were also incorporated into a re-assessment of the classic boundaries of the radiative habitable zone (Kopparapu et al., 2013). This climatic work was done in parallel with work that outlined the possible effects of various greenhouse gases (Byrne and Goldblatt, 2014), providing a library of greenhouse gas forcings applicable for studies of Earth and other planets (including early Mars).

We also re-considered whether the Archean atmosphere had an organic haze, and the implication for its climate. Our past geochemical work has suggested that a haze may have been present on Archean Earth (Zerkle et al., 2012; Kurzweil, 2013; Farquhar, 2013; Domagal-Goldman et al., 2009). We have computed the first self-consistent solutions for Archean Earth’s atmospheric chemistry and climate with a fractal hydrocarbon haze (Arney et al 2014a,b,c; Arney et al 2015). The climate model was updated to incorporate optical properties of hydrocarbon hazes, and coupled to the photochemical model. In addition, the coupled model haze outputs were used to generate planetary spectra using our SMART tools, allowing investigation of the detectability of hazy exoplanets. Our results are consistent with geochemical data and show that Archean Earth could have remained above 273 K even with a hydrocarbon haze and even if the surface pressure of the planet was less than 1 bar, as suggested by our team’s previous paleopressure proxies (Som et al., 2012).

The spectral effects of these hazes were also simulated, to facilitate the use of Archean Earth as an analog exoplanet (Arney et al., 2014). The hazes produced brighter planets at visible wavelengths, and a broad absorption feature in the UV that may be detectable at low spectral resolution. This UV feature also makes fractal hydrocarbon hazes a UV shield for the planet’s surface. We also generate transit transmission spectra of these hazy worlds and find they produce sloped transit transmission spectra similar to Titan’s (Robinson et al 2014). Because hazy exoplanets may be common, we investigated haze production for Archean Earthlike planets orbiting the modern sun, AD Leo, a modeled quiescent M dwarf, a K2V star, and an F2V star. We found that planets orbiting stars with very high UV flux (AD Leo, the F2V dwarf) or very low UV flux (the quiescent M dwarf) are unlikely to form significant hydrocarbon hazes in their atmospheres. The high UV planets destroyed hydrocarbons before they formed hazes, and the low UV planets could not dissociate sufficient quantities of methane to initiate higher order hydrocarbon formation. This was part of a broader effort to learn lessons from Archean Earth for future exoplanet imaging missions (Domagal-Goldman, et al., 2013).

Finally, we performed an analysis of isotope data to explore the transition from a low-O2 planet to one with significant atmospheric O2. Understanding what controlled the build-up of atmospheric O2 on our own planet is critical for understanding the possibilities for biosignatures on other worlds. Work was completed on examining constraints on O2 fluxes that can be deduced from Earth’s record of isotopes of sedimentary carbon (Krissansen-Totton et al., 2015; Krissansen-Totton and Catling, 2014). Permanent burial of organic carbon oxidizes the atmosphere and ocean by removing reduced carbon, which is typically enriched in 12C and so leaves seawater carbon, recorded in carbonates, depleted in 12C. We examined a conventional mass-balance model where global carbon input to surficial reservoirs is balanced by burial of sedimentary carbonates and organic carbon. Our statistical analysis showed that fractional organic burial has increased since 3.6 Ga by a factor of 1.2 to 2.0 with 95% confidence. An increase in organic burial by a factor of 1.2 cannot explain the 2.4 Ga rise of O2, whereas an increase by a factor of 2 could. We also found a statistically significant increase in the difference between carbonate and organic carbon isotopic ratios from the early to late Archean. A possible explanation is that the evolution of oxygenic photosynthesis at 2.8 Ga or earlier allowed sulfate or O2 to be used in methanotrophic oxidation of methane, which causes a large isotopic fractionation of carbon. Finally, we explored a more complex model with additional fluxes of ‘authigenic carbonates’, originating from oxidized organic matter, and ocean crust carbonatization. Because the size of these fluxes has a large, poorly constrained range, analysis including them showed that the carbon isotope record would not constrain the history of organic burial (Figure 2). However, the magnitude of these additional processes may have been inconsequential throughout geologic time, in which case conclusions from the conventional model would be valid.

Work was also done considering other hypotheses for what caused the rise of oxygen at 2.4 Ga. Geochemical evidence suggests that 0.2-0.7 b.y. elapsed between the advent of oxygenic photosynthesis and the rise of oxygen. Zahnle and Catling (2014) considered how that interval could have been determined by the rate of hydrogen escape to space and by the size of the reduced reservoir that needed to be oxidized before O2 became favored. They also discussed how hydrogen escape may have been linked to the history of continental growth.

Finally, we considered some constraints on O2 prior to the rise of oxygen in two special papers that commented on suggestions by other authors that O2 was higher than previously thought (Domagal-Goldman, 2014a) or lower than previously thought (Domagal-Goldman, 2014b). Both of these papers discuss the geochemical constraints that make these alternative hypotheses unlikely, but that discuss the possibilities for certain parts of the Earth system – the ocean in particular – to demonstrate redox states that do not directly track that of the atmospheric proxies at the time.

Example Suite of Photochemical Model Runs From Claire Et Al., (2014).
In this experiment, the total volcanic sulfur flux was increased while holding all other parameters constant. The upper panel shows mixing ratios and fluxes, and the lower panel shows predicted Δ33S and relative atmospheric exit channel proportions, as a function of increasing total volcanic sulfur flux.

Organic Burial Fraction as a Function of Time.
The dot-dashed line and grey shaded region (95% confidence interval) show the organic burial fraction, forg, since 3.6 Ga calculated using the conventional mass balance model. The solid line and dashed lines (95% confidence interval) show f_org from a more complex mass balance model that incorporates authigenic carbonates and ocean crust carbonatization. Both curves are derived from the same organic and carbonate isotope time series data using LOWESS (LOcally WEighted Scatterplot Smoothing) (Krissansen-Totton and Catling, 2014).

    Roger Buick Roger Buick
    Project Investigator
    David Catling David Catling
    Project Investigator
    James Kasting James Kasting
    Kevin Zahnle Kevin Zahnle
    Mark Claire Mark Claire
    Sanjoy Som Sanjoy Som
    Shawn Domagal-Goldman
    Project Investigator

    Giada Arney

    Benjamin Charnay

    Colin Goldblatt

    Eva Stüeken

    Objective 1.1
    Formation and evolution of habitable planets.

    Objective 1.2
    Indirect and direct astronomical observations of extrasolar habitable planets.

    Objective 4.1
    Earth's early biosphere.

    Objective 4.2
    Production of complex life.

    Objective 5.1
    Environment-dependent, molecular evolution in microorganisms

    Objective 5.2
    Co-evolution of microbial communities

    Objective 6.1
    Effects of environmental changes on microbial ecosystems