Global warming at near-constant tropospheric relative humidity is supported by observations

Climate models agree that global mean total precipitable water increases by about 7% per 1°C of global warming. New observations allow us to further support this and to narrow the range in global projections of tropospheric humidity in line with formerly constrained projections of global warming.
Published in Earth & Environment
Like

Water vapour is a major component of the Earth’s atmosphere and plays a key role in the planet’s energy balance. It is the main greenhouse gas (GHG) in the atmosphere that accounts for about 50% of the total greenhouse effect globally (against ~25% for clouds and ~20% for carbon dioxide). It is also an integral part of the water cycle and fuels precipitation events as it condensates under the effect of atmospheric updrafts. While precipitation averaged globally is controlled by the atmospheric energy budget, local extremes are primarily driven by thermodynamical and dynamical factors.

 

Based on the Clausius-Clapeyron equation established in the early 19th century, the water-holding capacity of the atmosphere increases by about 7% for every 1°C rise in temperature. It is often assumed that this rate of increase also applies to actual atmospheric humidity in a warmer climate, with expected first-order consequences on the overall water cycle intensity including extremes. In other words, and despite the lack of theoretical evidence, global warming is generally considered to occur at a near-constant tropospheric relative humidity. Yet, and so far, the observational evidence is still blurred by the large spatio-temporal variability of atmospheric humidity and the difficulty to provide long enough and quality-checked global datasets.

 

In-situ based humidity observations from radiosondes suffer from uncertainties over poorly sampled regions, particularly in the Southern Hemisphere. Satellite observations allow a global monitoring of water vapour, but also suffer from a number of limitations, especially over land. Data assimilation within global atmospheric re-analyses allows the fusion of in-situ and satellite observations with numerical weather predictions, but also suffer from strong inhomogeneities as long as the observations have not been quality-checked and/or the global observing system is changing over time. In this respect, the World Climate Research Program (WCRP) has provided various recommendations including the need for a bias-corrected multi-station radiosonde archive and of a homogeneous reprocessing of satellite data records.  

 

Such limitations explain the careful conclusions of the latest Assessment Report of the Intergovernmental Panel on Climate Change (IPCC, 2021). Chapter 2 assessed that positive trends in global total column water vapour are very likely since 1979 when direct observations began, although uncertainties associated with changes in the observing system imply medium confidence in estimation of the trend magnitudes. Similarly, Chapter 3 assessed that human influence has likely contributed to a moistening in the upper troposphere since 1979 and, with medium confidence, that it has also contributed to an increase in annual surface specific humidity over the mid-latitude Northern Hemisphere continents, but did not quantify the human contribution to the observed changes in the overall tropospheric humidity.

 

Our study makes use of a Bayesian statistical method (Kriging for Climate Change, Ribes et al., 2021), a recent reconstruction of the global mean total precipitable water (GTPW), and a longer and widely used record of global mean surface temperature to constrain simulated changes in atmospheric moisture from two generations of global climate models (CMIP5 and CMIP6 respectively). Our GTPW reconstruction is based on total column water vapour retrievals from a global network of ground-based Global Navigation Satellite System (GNSS, Bock 2022) stations. The KCC method assumes that “models are statistically indistinguishable from the truth” and allows us to derive a posterior distribution of simulated changes in GTPW from the raw model outputs (our empirical prior distribution) conditional to the available set of observations. Both past and future changes are constrained in a consistent way, and the contribution of each anthropogenic forcing (GHG but also other agents such as aerosols) can be diagnosed.

 

Our results show that past changes in atmospheric humidity were driven by radiative forcings from both GHG emissions and aerosols, with the latter damping the atmospheric moistening caused by the former over much of the 20th century. In contrast, the early 21st century increase in GTPW was dominated by the GHG emissions given the reversal of the trend in the aerosol radiative forcing (Quaas et al., 2022). Most CMIP6 models show a too strong atmospheric moistening compared to our 1994-2021 reconstruction. Our analysis reveals that this mismatch is primarily due to the overestimation of the concurrent observed global warming (Ribes et al., 2021), rather than to errors in the response of relative humidity.

The method also allows us to constrain future changes in GTPW. A substantial narrowing in the range of the CMIP6 projections is obtained throughout the 21st century (by about 39% in 2100) when using the SSP2-4.5 intermediate emission scenario. Similar results are obtained with a high-emission scenario, both with CMIP5 or CMIP6 models. This is mainly the result of the global mean temperature constraint given the limited length of the GNSS record. As for the recent past, many CMIP6 models overestimate the future atmospheric moistening. Yet, all models are right in projecting a global warming at near-constant relative humidity.

Our study provides a way to better constrain the projections of the atmospheric moistening. Our conclusions fully support the long-standing assumption that the Clausius-Clapeyron rate also applies to the actual tropospheric water content (not only its holding capacity). Note that there is however no theoretical reason why the rate of increase should be exactly 7%/°C at the global scale. Deviations from the Clausius-Clapeyron scaling of zonal-mean changes in precipitable water have been shown to result from decreases in relative humidity in the subtropics and mid-latitudes, and increases in the deep tropics. Further investigation is therefore needed to better constrain the 3D distribution of the water vapour response. This could be also useful to narrow residual uncertainties in the positive water vapour feedback (1.77 ± 0.20 W/m2/°C across CMIP5 and CMIP6 models).

 

Our findings also have some implications for adaptation. They further support an overall intensification of the global water cycle. The prevalent increase in atmospheric water vapour strengthens the atmospheric moisture transport and alters the surface and atmospheric energy balance, thereby influencing global evaporation and precipitation changes. Moreover, the water vapour latent heat fuels intense storms and its increase may then result in more extreme weather and precipitation events. While the human influence on such water cycle alterations has been documented by multiple event-attribution studies, a more comprehensive and accurate assessment of projected changes would benefit to the design of careful adaptation strategies.

References:

 

Bock, O. (2022). Global GNSS Integrated Water Vapour data, 1994-2021 [Data set]. AERIS. https://doi.org/10.25326/68

IPCC, 2021: Climate Change 2021: The Physical Science Basis. Contribution of Working Group I to the Sixth Assessment Report of the Intergovernmental Panel on Climate Change[Masson-Delmotte, V., P. Zhai, A. Pirani, S.L. Connors, C. Péan, S. Berger, N. Caud, Y. Chen, L. Goldfarb, M.I. Gomis, M. Huang, K. Leitzell, E. Lonnoy, J.B.R. Matthews, T.K. Maycock, T. Waterfield, O. Yelekçi, R. Yu, and B. Zhou (eds.)]. Cambridge University Press, Cambridge, United Kingdom and New York, NY, USA, In press, doi:10.1017/9781009157896.

Quaas, J., Jia, H., Smith, C., Albright, A. L., Aas, W., Bellouin, N., Boucher, O., Doutriaux-Boucher, M., Forster, P. M., Grosvenor, D., Jenkins, S., Klimont, Z., Loeb, N. G., Ma, X., Naik, V., Paulot, F., Stier, P., Wild, M., Myhre, G., and Schulz, M.2022: Robust evidence for reversal of the trend in aerosol effective climate forcing, Atmos. Chem. Phys., 22, 12221–12239, https://doi.org/10.5194/acp-22-12221-2022.

 

Ribes A, Qasmi S, Gillett N., 2021: Making climate projections conditional on historical observations. Sc. Adv., 7, eabc0671.

 

Please sign in or register for FREE

If you are a registered user on Research Communities by Springer Nature, please sign in