Back

Quantifying human contribution to terrestrial drying

Forcing agents graph (Download Image)

For the study, researchers considered soil moisture changes from 1971-2016 in response to six sets of forcing agents. The shaded regions represent the range of soil moisture changes arising from natural climate variability alone. Represented in this example are the observed signals for the month of October (represented by the black vertical lines), which can be explained by a combination of human and natural forcings, but cannot arise from natural climate variability alone. Diagram key: ALL (black solid, forced by all anthropogenic and natural forcing agents), ANT (brown dashed, forced by anthropogenic forcings only), GHGAER (purple dashed, forced by anthropogenic greenhouse gases and aerosols only), GHG (magenta solid, forced by anthropogenic greenhouse gases only), AER (blue solid, forced by anthropogenic aerosols only), NAT (green dotted, forced by natural solar and volcanic forcings only).

Historical drying trends have been demonstrated to occur over the lands surface, mostly in the subtropics and midlatitudes. Such drying trends are also widely projected to continue during the twenty-first century, especially under high greenhouse gas emission pathways.

The causes of terrestrial drying can be understood in terms of the effects of natural climate variability patterns, natural solar variability and volcanic eruptions, and human-induced greenhouse gases and aerosol emissions. Previous studies have effectively separated the causes of terrestrial drying by using formal detection and attribution (D&A) methods based on Earth system model (ESM) predictions with different combinations of forcing agents turned on and off. However, these studies focused on either meteorological drought, agricultural drought, or surface aridity indicators that did not reveal seasonal variations or vertical differences in moisture across soil layers.

To fill this research gap, LLNL scientist Celine Bonfils and collaborators at Oak Ridge National Laboratory used a Standardized Soil Moisture Index (SSI)—calculated from multi-source merged data sets—and a variety of simulations from the Coupled Model Intercomparison Project Phase 6 (CMIP6) to detect and attribute the changes in latitudinal terrestrial soil moisture between 1971–2016. For this, a pattern-based D&A method previously developed at LLNL was applied separately for each month of the year and each soil layer. The merged data sets use a comprehensive list of observations, reanalysis, and offline model simulations to reduce the potential biases and uncertainty caused by the limited sampling of source data.

The research team found greenhouse gases contributed significantly to the decrease in surface soil moisture (0–10 cm) from August to November and in root-zone soil moisture (0–100 cm) from September to April, with more rapid drying in 0–10 cm than 0–100 cm. Their findings demonstrate clear vertical, zonal, and seasonal patterns and highlight predominant human contributions to terrestrial drying, providing a basis for drought and flood risk management.

[Y. Wang, J Mao, F.M. Hoffman, C. Bonfils, H. Douville, M. Jin, P.E. Thornton, D.M. Ricciuto, X. Shi, H. Chen, S.D. Wullschleger, Quantification of human contribution to soil moisture-based terrestrial aridityNature Communications (2022), doi: 10.1038/s41467-022-34071-5.]