Modelling the long-term impacts of artificial warming on the Martian water cycle and surface ice distribution

Abstract

Recent papers by Ansari et al. (2024, Science Advances 10, eadn4650) and Richardson et al. (2025, arXiv eprint 2504.01455) have suggested that global warming of the Martian surface ('terraforming') by 35 K to sustain local habitats above the melting point of water could be achieved through the injection of engineered aerosols into the Martian atmosphere. Using the MarsWRF 3D Global Climate Model, we investigate how artificial warming of Mars through engineered aerosol release would affect the planetary water cycle and the distribution of the major surface ice reservoirs. Within our model framework, every 20 K of global warming induces a tenfold increase in atmospheric water vapour content due to sublimation of H2O ice from the North Polar Cap. This increases the potency of cloud radiative feedbacks which induces nighttime warming (~5-10 K) at low latitudes, but daytime cooling (up to 40 K) in the winter midlatitudes. Water is transferred from the edge of the North Polar Cap to the South Polar Cap and there is minor destabilisation of shallow northern midlatitude subsurface ice. As a result, seasonal sublimation of H2O ice from the South Pole has an increased impact on the global water cycle. These changes persist on Mars at least decades after loading of the atmosphere with engineered aerosols ceases. Our model is limited by the gaps in our knowledge of present-day Martian weather and climate, and of the microphysics and radiative properties of candidate warming agents. Much more data is therefore needed before warming Mars could become feasible.

Figures (18)

• Figure 1: Top row: a) simulated insolation reaching the surface and b) simulated emission at the top of the atmosphere (multiplied by a factor of 20) for both clear and dusty conditions with optical depth LGR τ = 3, excluding the effect of engineered aerosols. The vertical black dashed line represents the 15 LGR μm absorption band, where CO2 absorbs efficiently kutepov2025. An engineered aerosol absorbing at 15 LGR μm will have relatively little additional warming effect. c)-f): idealised Mars-warming aerosol optical properties, chosen to forward-scatter solar radiation to the surface while blocking surface thermal emission. We also show the spectrum of Martian dust with an effective radius of 1.5 LGR μm wolff2009haberle2019amesmodel for comparison. Natural Mars dust lowers Mars dayside temperatures Streeter2020, and is therefore not a suitable warming agent.
• Figure 2: Evolution over time of, from top to bottom: global average surface temperature, atmospheric water vapour column depth, atmospheric water ice column depth and aerosol column depth. Results are shown as a function of (left) engineered aerosol release rate for a fixed dry deposition velocity of 0.03 cm/s, and (right) dry deposition velocity for a fixed aerosol release rate of 2.5 l/s. The model runs shown in this figure all correspond to scenarios with radiatively active water vapour but non-radiatively active clouds
• Figure 3: Evolution of global average surface temperature, atmospheric water vapour and ice column, and aerosol column depth for a 2.5 l/s aerosol release rate scenario. Results are shown for three different radiatively active scenarios. The two scenarios with non-radiatively active clouds are mostly indistinguishable from each other as the greenhouse effect from water vapour is weak.
• Figure 4: Seasonal variations in (a) surface temperature and (b,d) the atmospheric water cycle following 10 Mars years of aerosol release. (c) compares diurnally averaged temperature increase due to the direct versus indirect effect of radiative cloud feedbacks in the model.
• Figure 5: Zonal average meridional cross-sections 10 Mars years after start of aerosol release. Radiatively active ("ra") clouds warm the middle atmosphere, allowing for water vapour to rise higher in the atmosphere before condensing. Cloud belts over low- and midlatitudes therefore warm Mars during the night, while low-altitude clouds over the winter midlatitudes cause dramatic daytime cooling. Recession of north polar CO2 ice cap results in strong water ice sublimation in Northern summer, and hence strong warming due to radiative cloud feedbacks.