TL;DR
This study analyzes historical and projected extreme precipitation trends across China using observational data and CMIP5 climate models. It reveals a general increase in extreme precipitation over the past 58 years and projects further intensification in the future, with significant regional variations, highlighting potential impacts on socio-economic and ecological systems.
What Problem Does This Paper Address?
Extreme precipitation events are becoming more frequent and intense globally due to climate change, posing significant risks to societies and ecosystems. While the general trend of increasing extreme precipitation is understood, the specific spatiotemporal patterns and future evolution within large, geographically diverse regions like China remain less clear. Previous studies have often focused on smaller scales or lacked comprehensive analysis of both historical observations and future projections. This paper aims to fill this gap by providing a detailed assessment of extreme precipitation variability in China over the past 58 years and projecting its changes for the next 90 years, considering regional differences and the impact of global warming.
Methods
The study employs a multi-faceted approach to analyze extreme precipitation. For historical data (1960-2017), they utilize observational data from the China Meteorological Administration (CMA). To characterize extreme precipitation, they calculate eight Extreme Precipitation Indices (EPIs) proposed by the World Meteorological Organization. Crucially, they use both fixed thresholds (e.g., 10mm, 20mm, 50mm daily rainfall) and a percentile-based threshold (95th percentile) to define extreme events. The percentile approach is highlighted as a way to account for geographic variations in rainfall.
For future projections (2011-2100), they leverage data from five Coupled Model Intercomparison Project Phase 5 (CMIP5) models under the Representative Concentration Pathway (RCP) 8.5 scenario, which represents a high greenhouse gas emission pathway. To select the most reliable model for China, they compare the simulations of historical data with observations, choosing the BCC-CSM1-1 model based on its highest correlation and lowest RMSE. The projected data is then systematically corrected to account for model biases. The analysis involves fitting linear trends to the EPIs using ordinary least squares regression and assessing the statistical significance of these trends with the Mann-Kendall test.
Datasets
- Observational Data: Daily precipitation data from the China Meteorological Administration (CMA) meteorological stations (1960-2017). This dataset includes historical observations from 824 sites.
- Climate Model Projections: Data from five CMIP5 models: BCC-CSM-1-1, BNU-ESM, FGOALS-g2, GFDL-ESM2M, and NorESM1-M. Projections cover the period 2011-2100 under the RCP 8.5 scenario. Historical simulations (1960-2005) from these models were also used for model selection and bias correction.
Key Insights and Ideas
- North-South Gradient: A consistent spatial pattern emerges where extreme precipitation indices generally decrease from south to north across China, mirroring the country's diverse topography and climate zones.
- Regional Divergence in Trends: While the overall trend for China is an increase in extreme precipitation, specific regions exhibit contrasting patterns. For instance, the northwest, southeast, and Tibet Plateau saw increases in the past, while northern China experienced decreases.
- Future Intensification: Projections indicate a continued increase in extreme precipitation across China in the coming decades, though the magnitude of future events is expected to be lower than the historical peak values.
- Model Selection is Crucial: The study emphasizes the importance of evaluating and selecting appropriate climate models for regional climate change studies, as different models exhibit varying degrees of accuracy in simulating historical extreme precipitation patterns.
- Percentile Thresholds Matter: Using a 95th percentile threshold, in addition to fixed thresholds, helps to better capture extreme events across diverse geographical locations with varying baseline precipitation levels.
Key Results
- Historical Trends (1960-2017):
- Overall increase in most EPIs across China.
- Significant upward trends observed for R10, R20, R50, and R95d.
- Extreme precipitation increased in the northwest, southeast, and Tibet Plateau, but decreased in northern China.
- Spatial patterns of EPIs generally showed a decrease from south to north.
- Projected Trends (2011-2100):
- All EPIs are projected to increase in the future, with the average values in the second half of the century (2056-2100) generally higher than the first half (2011-2055).
- The SE and SW regions are expected to experience more severe extreme precipitation events.
- An exception is the intensity of extreme precipitation (AEPI), which is projected to decrease in southeastern China during the second period (2056-2100).
- The trends in the second period (2056-2100) are generally expected to be smaller than in the first period (2011-2055), suggesting a slower rate of increase.
Limitations and Future Work
- Model Uncertainties: Despite selecting an optimal model, projections still carry uncertainties, particularly under the high-emission RCP 8.5 scenario. The use of only one RCP scenario might also limit the scope of future projections.
- Regional Differences in Model Performance: While the BCC-CSM1-1 model was chosen as optimal, the study notes that CMIP5 models generally struggle to perfectly replicate historical EPIs, with some overestimating and others underestimating certain indices.
- Limited Characterization of Event Duration: The study acknowledges that while eight EPIs characterize basic features of extreme precipitation, they do not fully capture the duration of these events.
- Future Work: The authors suggest that future research should focus on the impacts of extreme precipitation on ecosystems and human life, and potentially explore non-significant trends more deeply. Further investigation into the duration of extreme precipitation events is also warranted.
Summary updated