Attribution of Summer Total Precipitation Changes in China: Dominant Role of Circulation Pattern Frequency Driven by Greenhouse Gas Forcing

Summer total precipitation changes over eastern China between 1961 and 2014 were overwhelmingly driven by shifts in synoptic circulation pattern frequency rather than thermodynamic effects alone, according to a study published on AGU Publications.

Detecting and attributing long-term precipitation variations across China has historically proved difficult due to pronounced spatial heterogeneity and generally poor model performance. The observed summer total precipitation since the 1960s showed an overall trend of wetness in the south and dryness in the north. Different outside forces overlapped in various subregions, making it hard to reach firm conclusions.

Researchers proposed a novel attribution framework to decompose precipitation changes into three multiplicative terms. These terms encompass the frequency for a synoptic circulation pattern representing dynamic effects, the frequency of precipitation under that pattern, and the mean intensity of precipitation when it occurs representing thermodynamic effects.

To implement this framework, daily boreal summer rainfall for the period 1961–2014 was collected at 753 meteorological stations across the China Meteorological Administration network. After quality control, a total of 674 stations with more than 24 years of record were actually used for the study.

Daily atmospheric circulation data for the same period came from the ERA5 reanalysis data set provided by the European Center for Medium-Range Weather Forecasts, which features a 0.25° × 0.25° resolution. Variables comprised geopotential height, wind, relative humidity, temperature on pressure levels, and sea level pressure.

A nonhomogeneous hidden Markov model was adopted to classify daily summer precipitation over China in a supervised manner. This approach incorporated concurrent large-scale circulation backgrounds to yield more typical and non-overlapping summer patterns. The optimal number of rainfall patterns was determined to be six.

Greenhouse Gas Forcing and Circulation Patterns

The analysis revealed that changes in circulation pattern frequency played a dominant role in summer total precipitation changes across eastern China during the 1961–2014 period. These dominant changes are very likely attributable to anthropogenic greenhouse gas forcing.

Climate models demonstrated superior performance in reproducing these circulation pattern variations. That performance gave key confidence to the findings that human activity was responsible.

Previous studies using optimal fingerprinting methods found that human influence on China’s mean precipitation changes was detectable, but results remained at the limit of statistical significance. Process-based approaches previously pointed to the weakening of the Asian westerly jet caused by uneven anthropogenic aerosol emissions as a driver for dipolar precipitation trends in High Mountain Asia during 1951–2020.

Other analyses indicated that the north-south drying-wetting pattern across China was largely driven by precipitation frequency potentially due to aerosols. However, those prior findings lacked direct evidence because self-organizing map analyses resulted in overlapping precipitation patterns that prevented the separation of different external forcing signals.

Model Simulations and Experimental Data

Daily precipitation and circulation variables from CMIP6 GCMs were utilized for detection and attribution testing. Historical simulations included multiple runs under all external forcings, greenhouse gas-only forcings, anthropogenic aerosol-only forcings, natural-only forcings, and preindustrial control experiments.

Four models simultaneously provided daily precipitation and circulation variables with a minimum of one member per forcing experiment. Twelve models offered preindustrial control experiments.

Researchers used 95 non-overlapping 54-year segments corresponding to the length of the 1961–2014 period from the preindustrial control simulations. These control simulation years did not correspond to the actual historical years.

Established in 1975, the European Center for Medium-Range Weather Forecasts functions as an independent intergovernmental organisation supported by most of the nations of Europe. The Centre was created to pool scientific and technical resources for medium-range weather forecasting and operates one of the largest supercomputer complexes in Europe alongside the world’s largest archive of numerical weather prediction data.

The framework addressed the limitation of older methodologies that implicitly assumed rainfall occurred every time a circulation pattern appeared. By further decomposing the thermodynamic term into actual precipitation event probabilities and precipitation intensity, the study provided more physically constrained evidence for regional-scale attribution.

Photo of author

Technology Desk

Technology Desk at Archyde.

PNP: Ex-Duterte Staff Not Held Captive, No Rescue Conducted