Haibo Du, Lisa V. Alexander, Markus G. Donat, Tanya Lippmann, Arvind Srivastava, Jim Salinger, Andries Kruger, Gwangyong Choi, Hong S. He, Fumiaki Fujibe, Matilde Rusticucci, Banzragch Nandintsetseg, Rodrigo Manzanas, Shafiqur Rehman, Farhat Abbas, Panmao Zhai, Ibouraïma Yabi, Michael C. Stambaugh, Shengzhong Wang, Altangerel Batbold, Priscilla Teles de Oliveira, Muhammad Adrees, Wei Hou, Shengwei Zong, Claudio Moises Santos e Silva, Paulo Sergio Lucio, Zhengfang Wu
Geophysical Research Letters
Work Package 3
Link: https://doi.org/10.1029/2019GL081898
Highlights
Understanding trends in extreme rainfall events are an important part of managing and mitigating the threat they pose. This study uses a new compilation of rainfall observations in combination with climate model simulations to study the trend in rainfall events over their entire duration, termed persistent rainfall, rather than simply studying daily extremes. This aggregate amount over an event’s duration may be more relevant in terms of impacts of the event. The team found that global rainfall aggregated over the entire duration of events is, on average, higher than the daily rainfall amounts. The trend in persistent rainfall is often smaller than daily rainfall, though with a great deal of regional variation, but they still show an increase in extreme rainfall in the recent past and in the future. Demonstrating the importance of persistent rainfall in the analysis of extreme rainfall events will allow better predictions of extreme events to be made, allowing people and authorities to put more effective plans in place to mitigate their effects.
Background
Extreme weather events, including extreme rainfall, can be extremely damaging, and studying how they may change in future is essential to mitigating their impacts. The intensity of extreme daily rainfall is expected to increase as the climate changes, both on a global and regional scale, however most of these studies focus on rainfall within a specific time period, for instance one day. There is a distinct lack of studies that assess extreme rainfall events over their entire duration, which may be across multiple days, termed persistent rainfall. Such studies would capture complexities in extreme rainfall patterns that are overlooked by fixed-duration studies, as well as assessing the influence of climate change on rainfall events over their entire duration. This study attempts to address this deficiency, using a new compilation of rainfall data to overcome the lack of data coverage that has previously held back studies of persistent extreme rainfall.
Results
The team found that global persistent rainfall is, on average, higher than the daily rainfall amounts. Daily rainfall measurements may divide a single, long event into multiple single-day rainfall events, therefore persistent rainfall measurements were better able to represent real events and trends. The global persistent rainfall trends show lower relative increases than the daily trends, both in observations and models, primarily influenced by slower increases in persistent rainfall than for daily rainfall in areas south of 45°N latitude. These spatial differences extend to some areas showing opposite changes in persistent and daily extreme rainfall. Both persistent and daily extreme rainfall are predicted to increase as the climate warms, with the magnitude of the changes increasing with additional warming. Uncertainties in the projections of future trends are higher in the tropics and sub-tropics, while some regions, again, show differing predicted changes between persistent and daily extreme rainfall, some with an increasing trend and others with a decreasing trend. Further work is needed to shed light on these uncertainties and discrepancies, but this study does show that persistent extreme rainfall analysis provides important insight into how extreme rainfall events are changing in our warming world.
Methods
This study makes use of a large compilation of observed global rainfall data gathered from 15 countries, most of which has been previously unavailable for global precipitation studies. After quality control tests, the dataset included data from 6,125 stations. Past trends in extreme rainfall were studied for the period from 1961 to 2010. Future predictions of extreme rainfall patterns up to 2100 were made using 30 models that contributed to the CMIP5 project, with two possible scenarios of future warming, moderate (RCP4.5) and high (RCP8.5).
Policy relevance
Improving predictions of extreme rainfall is crucial to limiting the impacts of these extreme events on society. The improved understanding of extreme rainfall trends this study brings could help people, businesses and authorities put better policies in place to adapt to and mitigate the impacts of extreme weather events, such as flooding.
Abstract
Extreme precipitation often persists for multiple days with variable duration but has usually been examined at fixed duration. Here we show that considering extreme persistent precipitation by complete event with variable duration, rather than a fixed temporal period, is a necessary metric to account for the complexity of changing precipitation. Observed global mean annual‐maximum precipitation is significantly stronger (49.5%) for persistent extremes than daily extremes. However, both globally observed and modeled rates of relative increases are lower for persistent extremes compared to daily extremes, especially for Southern Hemisphere and large regions in the 0‐45°N latitude band. Climate models also show significant differences in the magnitude and partly even the sign of local mean changes between daily and persistent extremes in global warming projections. Changes in extreme precipitation therefore are more complex than previously reported, and extreme precipitation events with varying duration should be taken into account for future climate change assessments.
Funded by the European Union under Horizon 2020.