Een mooi artikel van mei 2024 op nature. Het is een lang artikel maar goed onderbouwd. Sommige stukken met formules gaan mijn pet ver te boven.

Volledige artikel op onderstaande link.

https://www.nature.com/articles/s43247-024-01442-3

Abrupt reduction in shipping emission as an inadvertent geoengineering termination shock produces substantial radiative warming

Abstract

Human activities affect the Earth’s climate through modifying the composition of the atmosphere, which then creates radiative forcing that drives climate change. The warming effect of anthropogenic greenhouse gases has been partially balanced by the cooling effect of anthropogenic aerosols. In 2020, fuel regulations abruptly reduced the emission of sulfur dioxide from international shipping by about 80% and created an inadvertent geoengineering termination shock with global impact. Here we estimate the regulation leads to a radiative forcing of +0.2 +-0.11Wm $+ 0.2 \pm 0.11$

Introduction

The Earth’s atmosphere has warmed because of human activities increasing the concentration of greenhouse gasses that trap thermal radiative energy in the climate system, creating a positive climate forcing. Human activities have also increased the concentration of aerosol particles that can affect the amount of reflected solar radiation back to space either directly or indirectly by interacting with clouds, which has an overall cooling effect on the climate1. The magnitude of the aerosol cooling effect has significant implications for estimating how sensitive our climate is to greenhouse gas forcing and the amount of expected future warming for a given increase of greenhouse gas concentrations2. The effectiveness of anthropogenic aerosols in cooling the climate also has direct implications for solar radiation modification geoengineering schemes3,4. Such methods aim to produce temporary cooling of the climate through enhanced reflection of solar radiation to space. They are not solutions to greenhouse gas induced global warming and have uncertain and complex additional consequences besides the intended short-term cooling effect4,5,6,7.

Marine cloud brightening (MCB) is a type of solar radiation modification scheme where marine low clouds are seeded with aerosols to become brighter8,9. Examples of small scale, opportunistic MCB experiments were discovered in early satellite observations of ship-tracks, linear features of brighter oceanic clouds because of ship-emitted aerosols10,11. The addition of aerosols from ship emissions results in more cloud droplets, leading to more reflective clouds for a given amount of total In-cloud liquid water, or liquid water path (LWP)12. More recent studies show that aerosols can also change LWP and total cloud fraction (CF), which also greatly affect the amount of solar radiation reflected by clouds2,13,14,15,16.

Aerosols sourced from global shipping industry affect clouds and we can view the shipping emission as a long-running inadvertent MCB experiment. On January 1, 2020, new International Maritime Organization (IMO) regulations on the sulfur content of international shipping fuel took effect. The IMO 2020 regulation (IMO2020) reduced the maximum sulfur content from 3.5% to 0.5%17. While IMO2020 is intended to benefit public health by decreasing aerosol loading, this decrease in aerosols can temporarily accelerate global warming by dimming clouds across the global oceans. IMO2020 took effect in a short period of time and likely has global impact. IMO2020 effectively represents a termination shock for the inadvertent geoengineering experiment through a reverse MCB, i.e., marine cloud dimming through reducing cloud droplet number concentration (Nd) (Fig. 1). Observations of ship-tracks suggest that IMO2020 has reduced the occurrence and modified the properties of ship-tracks across global oceans, demonstrating that a regulation intended to reduce pollution had collateral effects on cloud microphysics18. Analyses of remote sensing data have shown evidence of cloud dimming in the South Atlantic shipping lane19. Outside the South Atlantic, the effect of IMO2020 does not have a distinct spatial structure18,19, which makes direct observation of the impact more challenging.

Results

Here we combine satellite observations and a chemical transport model to quantify the radiative forcing of the inadvertent geoengineering event induced by IMO2020 and estimate its climate impacts. We simulate the impact of IMO2020 on maritime aerosol concentrations with the NASA GEOS-GOCART model. Figure 1 shows the modeled reduction in aerosol optical depth due to decreased SO2 emission from the international shipping industry. The AOD reduction reaches peak values of around 0.01 in the South China Sea and Eastern North Atlantic off the coasts of Western Europe. In the South Atlantic the regulations create AOD reductions that follow the shape of shipping routes. We then calculate the ratio between the AOD change due to IMO2020 and that between pre-industrial and present day. Over most of the ocean, the ratio is smaller than 10% because of sparse shipping outside the major shipping routes. Over the North Pacific and North Atlantic, on the other hand, it can exceed 10% and reaches 25% in the Norwegian Sea and off the western European and northwestern African coasts. In these regions, the total anthropogenic aerosol concentration is relatively low because of declining emissions of aerosols and their precursors since the 1980s, making ship-emitted aerosols an important component of the anthropogenic maritime aerosols. The IMO2020 is therefore effective in reducing total aerosol loading for these regions. The impact of IMO2020 on the cloud droplet number concentration (Nd) of low-level maritime clouds as shown in Fig. 1C (see Methods). Globally, IMO2020 leads to a modest reduction of 0.5 cm−3 in mean modeled Nd. Regionally, however, the reduction is more pronounced. The strongest reduction occurs in the North Atlantic, the Caribbeans and the South China Sea, reaching 3 cm−3. These are regions with the busiest shipping lanes and thus strongest reduction of ship emissions. The reduction in the South Atlantic shows the most well-defined shipping lane shapes likely due to the unique circulation pattern in this region18,20. Figure 1D shows the ratio between IMO2020 induced Nd decrease and estimated Nd difference between preindustrial and present day. The ratio is small over the major outflow areas downwind of major continental sources, but becomes substantially larger in more remote oceans, reaching 30%. In the tropical North Atlantic, IMO2020 induced change in Nd can be more than 50% of the total anthropogenic change.

We combine Nd changes due to IMO2020 with satellite observations to estimate the forcing introduced by the inadvertent geoengineering event21. We consider both the Twomey effect and the effects of cloud liquid water path (LWP) and cloud fraction adjustments to Nd (see Methods section). The LWP and cloud fraction adjustments follow the functional forms derived from a large sample of ship-tracks21 that depend on the background cloud Nd, sea surface temperature (SST), estimated inversion strength, and background low cloud fraction (see Methods). Figure 2 shows the pattern of annual mean of forcing resulting from Nd decrease due to IMO2020 averaged over different LWP and cloud fraction adjustment functional forms. The total forcing is +0.2 +-0.11Wm² averaged over the global ocean with the Twomey effect contributing 40%, the LWP adjustment being near neutral, and the cloud fraction adjustment contributing 60%. The positive radiative forcing has strong regional variations. The North Atlantic experiences the strongest radiative forcing peaking around 1.4Wm−2 and whose basin-wide mean is around 0.56Wm−2. Weaker but still notable radiative forcing is seen in the North Pacific and the South Atlantic. This ordering is consistent with the amount of ship traffic and low cloud fraction in these regions. Our estimate of radiative forcing from IMO2020 is well within the range of estimates of the total forcing from shipping emissions in the literature22,23,24,25. We also compare our estimate with that from a recent observational study in the core shipping lane in the South Atlantic that used a different approach19. The two completely independent approaches yield very similar radiative forcing in the core shipping lane (supporting online material, SOM), which serves as a cross-validation. Similar global forcing, i.e., on the order of 0.1Wm−2, is reported by multiple modeling groups26.

$+ 0.2 \pm 0.11$

Using an energy balance model27, we calculate the expected amount of transient temperature increase due to warming resulting from IMO2020. For simplicity, we ignore the heat uptake by the deep ocean during the short-term, i.e. O(10) years. 0.2 W m−2 translates to around 0.16 K of warming with a timescale of 7 years. It is equivelant to 0.24 K/decade, which is more than double the average warming rate since 1880 and 20% higher than the mean warming rate since 1980, linear trend of 0.19 K/decade. We also calculated the lower and upper bounds of the forcing and corresponding expected warming (Fig. 3). The IMO2020 is expected to provide a substantial boost to the warming rate of global mean temperature in the 2020 s. The rate of warming is expected to ramp up quickly from 2020 and asymptotes to the longer-term trend line at the end of 202027. The 2023 record warmth is within the ranges of our expected trajectory. The magnitude of IMO2020 induced warming means that the observed strong warming in 2023 will be a new norm in the 2020 s. The mean temperature anomaly of the 2020 s will be 0.3 K higher than that of the 2010s. Regionally, the warming effect from IMO2020 on SST is harder to estimate since basin-wide SST changes can be affected by variations in factors like other aerosol concentration, ocean circulation, and air-sea interactions. However, the strong geographical variations in the forcing suggest the impact of IMO2020 on SST may have significant variation among ocean basins. For example, the North Atlantic SST may be disproportionally warmed more by the IMO2020 given the radiative forcing is more than three times the global average, which is likely a contributing factor to the pronounced warming of the North Atlantic SST in recent years28. A more robust quantitative estimate of the contribution of IMO2020 to regional SST warming requires coupled global climate models that have good representation of aerosol indirect effects.

The IMO2020-induced radiative forcing exhibits considerable seasonal variations. This is evident in the North Atlantic where the IMO2020 produced the strongest forcing. Figure 4 shows the monthly mean time series of forcing and its three components. We use a simple functional form for cloud adjustments that only depends on background Nd to illustrate the point. The total forcing varies between 0.19Wm−2 and 0.38Wm−2, a 100% relative change. The seasonal variation of incoming solar radiation is the dominant driver for this (SOM). But seasonal variations of background CF, Nd, and ΔNd due to IMO2020 also contribute as they affect the magnitude of the Twomey effect and macrophysical (LWP and CF) cloud adjustments. We estimate the contribution from each variable after removing the seasonal change in solar insolation, and report the results in Fig. 4B–D (see Methods). ΔNd induced by IMO2020 is the strongest contributor. Its variations can affect the forcing by more than 30% in some months such as Jan, Apr, and Dec. Its impact on LWP and CF adjustments contributes equally to the total radiative forcing. The seasonal variation of background Nd is also an important factor (Fig. 4C). Background CF also meaningfully contributes to the seasonal variations through mostly affecting the Twomey effect (Fig. 4D).

We compare the radiative forcing due to IMO2020 and its effect on radiative energy balance with observed changes in relevant quantities. The comparison does not prove causality but provides a context to assess the impact of IMO2020. The low cloud dimming forcing of 0.2 Wm−2 from the IMO2020 represents a strong temporary shock to the net planetary heat uptake (Fig. 5A) that has been increasing at a rate of ~0.05 Wm−2/yr29 in measurements. The net planetary heat uptake has increased by 0.25 Wm−2 since 2020, making the 0.2 Wm−2 due to IMO2020 nearly 80% of the total increase. The long-term trend of CERES TOA net radiation is 0.46 Wm−2/decade while it changes to 0.67 Wm−2/decade since IMO2020 took effect. The difference is 0.21 Wm−2 that is consistent with our estimated forcing. However, the record since 2020 is too short to ascertain the impact of IMO2020 on the long-term trend of the energy balance given its large interannual variations. The IMO2020 effect also has an asymmetric impact on aerosol loading in the northern and southern hemispheres because of higher baseline ship emissions in the northern hemisphere. This creates interhemispheric contrast in the resulting radiative forcing, which has important implications for deliberate geoengineering schemes because interhemispheric forcing contrast can create significant perturbations in precipitation patterns6. We calculate the interhemispheric contrast in IMO2020 induced warming effect to be around 0.22 Wm−2, with the northern hemisphere at 0.32 Wm−2 and the southern hemisphere at 0.1 Wm−2. The 0.22 Wm−2 contrast is substantial when compared with recent measured changes in the interhemispheric contrast in absorbed solar radiation. Figure 5B, C shows measured time series of top-of-atmosphere (TOA) absorbed solar radiation of both hemispheres and their contrast, respectively. Since IMO2020 took effect, the northern hemisphere (NH) absorbed solar radiation has increased by 0.5 Wm Quote selectie