Greenhouse gases have reshaped the atmospheric circulation patterns that steer summer rain across China, and that dynamic shift — not simply a warmer, wetter atmosphere — explains the dominant trend in seasonal precipitation since 1961.
Climate models have struggled to simulate China’s total summer rainfall, torn between the “southern flood, northern drought” dipole that masks a coherent human signal in regional totals.
“Changes in CP frequency that play a dominant role in precipitation changes over eastern China are successfully attributed to anthropogenic greenhouse gas forcing,” the authors wrote.
The team classified daily summer weather over China into six distinct rainfall patterns — circulation patterns, or CPs — using a nonhomogeneous hidden Markov model. The analysis drew on 674 meteorological stations with at least 24 years of records and ERA5 reanalysis data at 0.25° resolution.
Total precipitation change was then split into three multiplicative terms: the frequency of each circulation pattern, the probability of rain on a pattern day, and the mean rain intensity. The first term is dynamic; the latter two are thermodynamic. Across eastern China, circulation-frequency changes dominated the observed trend from 1961 to 2014.
To attribute the cause, observations were compared against CMIP6 simulations. Models driven by all forcings, and by greenhouse-gas-only forcings, reproduced the circulation-frequency shifts. Aerosol-only and natural-only runs did not. Internal variability was estimated from 95 non-overlapping 54-year segments of pre-industrial control runs.
“Models’ superior performance in reproducing changes in CP frequency lends confidence to our attribution findings,” the paper says.
Previous optimal-fingerprinting work on relative precipitation anomalies detected human influence only at the limit of statistical significance. Process-based studies had implicated aerosol-driven jet-stream changes or precipitation-frequency shifts but lacked direct attribution evidence. The NHMM approach yields non-overlapping patterns, which the authors say allows cleaner separation of forcing signals.
The framework offers a physically constrained path for regional attribution, though the authors note remaining uncertainties in thermodynamic contributions and model spread.
Related reading
- Alibaba Beats Revenue Estimates on 53% AI Cloud Surge, Warns of US Curbs Impact
- Researchers Predict Steady-State Entanglement in Diamond Qubits Using Synthetic Antiferromagnets
- Simon Harris’s Constituency Manager John Kelly Faces Four Complaints Since 2019 (archyworldys.com)
- Inflation Since October 2024 Adds About $4,720 a Year to Household Costs (daybreakwire.com)