Cirrus clouds and their geoengineering potential

Abstract

This thesis evaluates the option of modifying cirrus clouds in order to counteract part of the anthropogenic global warming, also known as cirrus seeding. The feasibility of cirrus seeding is assessed with the general circulation model ECHAM6-HAM with coupled aerosol-cloud interactions. The warming effect of cirrus clouds on climate led part of the research community to the idea of artificially decreasing their frequency and optical thickness by seeding them with ice nucleating particles. Cirrus seeding re- lies on the competition between two distinct cirrus formation pathways: homogeneous nucleation of soluble aerosols and heterogeneous ice crystal nucleation with the help of solid ice nucleating particles. Seeding attempts to transform optically thicker, longer lived homogeneously nucleated cirrus clouds to thinner and shorter lived heteroge- neously nucleated cirrus clouds. The cirrus seeding effectiveness depends on the correct representation of cirrus clouds in climate models. We evaluated the ECHAM6-HAM model against CALIPSO satellite data and found that the model reproduces the cirrus cloud occurrence fraction and ice water content well, but overestimates their extinction. Most importantly, we found that a large fraction of cirrus clouds formed in environments with liquid water, which cannot be modified by cirrus seeding. Moreover, modelled in situ cirrus formed by heterogenous ice nucleation on dust aerosols in most of the world, limiting the cir- rus geoengineering potential. Meanwhile, homogeneous cirrus were dominant only in the tropical tropopause layer and over mountain regions. Seeding cirrus clouds with ice nucleating particles of radii below 10 μm did not lead to a significant cooling effect on climate, due to radiative warming effects of increasing cir- rus cloud cover and decreasing ice crystal radius, which neutralized the cooling gained by the ice crystal number decrease. However, when seeding with larger ice nucleating particles, cirrus cloud cover, ice crystal number concentration, and ice water content decreased, due to increased ice crystal sedimentation velocity, leading to a cooling of up to 0.7◦ C globally. In addition, an idealized cirrus geoengineering setup with in- creased ice crystal sedimentation velocities at temperatures colder than -35◦ C could fully counteract the temperature increase by 1.5 x CO2 concentrations. Furthermore, seeding with ice nucleating particles larger than 10 μm could counteract up to 55% of the temperature and precipitation damage caused by a 1.5 x CO2 increase, without causing negative side effects in any of the analysed regions. Thus, cirrus geoengi- neering was found to be, not considering its problematic engineering details and large modelling uncertainties, an attractive method to counteract part of the anthropogenic warming.