The climate model is biased, and the constraint correction can better predict the summer precipitati

climate

"To deal with climate change, it is necessary to use climate models to predict future changes. The temperature overheating simulated by the latest climate models will bring deviations to the prediction results. The proposed constraint correction method is like linking the historical credit performance of 'credit card' holders with future rights and interests to ensure the best future returns, that is, accurate prediction results." Zhou Tianjun, a researcher at the Institute of Atmospheric Physics, Chinese Academy of Sciences This introduces the team's research results published in "Nature-Communications" on May 10.

By designing a constrained correction method with clear physical meaning based on observational data, the research team focused on the mid- and long-term changes in Asian-African monsoon precipitation. The prediction results of , have been constrained and corrected to overcome the bias caused by overheating, and significantly improve the reliability of the prediction results of summer precipitation in the Asian-African monsoon region.

According to their research, the regional average precipitation increase in the constrained Asian-African monsoon region is only about 70% of the original result. The new estimated data has reference value for the future water resources management of Asian and African monsoon countries.

New conundrum for international climate models: CMIP6 model overheating causes forecast bias

Adaptation to and mitigation of climate change requires the use of climate models to project future changes. The averaging of multiple climate model ensembles (MME) based on the "model democracy" principle of "one person, one vote" is a commonly used method in the scientific community. However, this approach faces challenges when applied to the latest climate models represented by CMIP6, which are sensitive to greenhouse gases due to uncertainties in the parameterization of complex processes in the climate system. Problems such as excessively high temperatures and overheating of simulated temperature changes can lead to an overestimation of the risk of future climate change impacts.

In order to solve this problem, the sixth assessment report of the Intergovernmental Panel on Climate Change (IPCC) proposed effective correction techniques for the prediction of global average temperature. However, there is no unified and effective method in the scientific community on how to solve the problem of projection bias caused by the overly high climate sensitivity of the CMIP6 model at the regional scale.

“Climate sensitivity is the magnitude of global temperature warming after a doubling of greenhouse gas concentrations. If climate models are more sensitive than they actually are, it makes projected future temperature changes 'too hot', overestimating the impact of climate change on including The impact of climate risks including extreme heat waves and extreme precipitation." Zhou Tianjun explained that for the Asian-African monsoon region, the impact is manifested in a significant overestimation of the future increase in summer precipitation and total runoff in the Asian-African monsoon region.

Effective solutions to overcome pattern bias

Zhou Tianjun's research team found that under two moderate greenhouse gas emission scenarios, SSP5-8.5 and SSP2-4.5, which represent high energy consumption and maintain the current level, respectively, the CMIP6 model has a significant impact on the mid- and long-term (2050-2099) summer precipitation in the Asian-African monsoon region. The estimated uncertainty of changes is mainly related to the simulation bias of the historical warming spatial distribution: the higher the climate sensitivity of the model, the stronger the warming rate in the northern hemisphere than in the southern hemisphere in the historical climate simulation, and the corresponding change in the Asian-African monsoon circulation is also higher. The stronger it is, the larger the projected change in future monsoon precipitation will be. Based on the observational data, the prediction results of summer precipitation were constrained by the inter-hemispheric thermal difference, and it was found that the original CMIP6 model significantly overestimated the increase of summer precipitation in the Asian-African monsoon region. Compared with the reference state (1965-2014), the future average precipitation in the Asian-African monsoon region (2050-2099) will increase by only 70% of the original result of the CMIP6 model, and the largest decrease will be in the West African monsoon region, which is about 49% of the original value.

Changes in precipitation also affect changes in runoff. Based on the relationship between precipitation and total runoff, the constrained precipitation forecast data was used to estimate the change in total runoff. The study found that the area with a significant increase in total runoff in the future accounted for only 66% of the original results. The most affected is the West African monsoon region.

or affect China's future water resources management

The Asian-African monsoon system is an important part of the global monsoon system, including the East Asian, South Asian and West African monsoons. The water resources for billions of people in the Asian-African monsoon region come from summer monsoon precipitation. Future changes in the Asian-African monsoon will have a significant impact on water resources and food production. Therefore, accurate prediction of precipitation changes in the Asian-African monsoon region is crucial for conducting climate change impact and adaptation assessments, and formulating mitigation strategies.

Research by Zhou Tianjun's research team shows that, due to the bias of the climate model in the warming spatial model, the original climate model overestimated the future changes in precipitation in the Asian-African monsoon region. "The constrained projections show that the increase in precipitation and total runoff in the Asian-African monsoon region is not as strong as the original results, but since the total runoff is directly related to the available water resources, it also raises new questions for water resources management, especially In the context of the observation that the precipitation in the monsoon region has shown a significant decreasing trend in the past few decades." Zhou Tianjun said.

Zhou Tianjun said that under two shared socio-economic paths representing "high carbon emission" and "moderate carbon emission" intensity, the projected future average precipitation in the Asian-African monsoon region (2050-2099) is compared with the reference state (1965). – 2014), which is 30% smaller than the original estimate; the largest decrease is in the West African monsoon region, which is about 49% of the original value; the estimated result in the East Asian monsoon region is about 24% smaller, It covers most of the area east of the densely populated Hu Huanyong Line in China.

Changes in precipitation also affect changes in runoff. Based on the link between precipitation and total runoff, the research team used the constrained precipitation forecast data to estimate the change in total runoff, and found that the area with a significant increase in total runoff in the future accounted for only 66% of the original results.

According to reports, the research work was jointly funded by the National Key Research and Development Project (2020YFA0608904, 2018YFA0606501) and the National Natural Science Foundation of China (41988101). The first author of the paper is Chen Ziming, a doctoral student at the University of Chinese Academy of Sciences, and the corresponding author is researcher Zhou Tianjun.

Fig. 1 Spatial cumulative probability density distribution (PDF) map of projected precipitation (a1~d1) and total runoff (a2~d2) changes under the SSP5-8.5 scenario. The colored part of the PDF and the number next to it indicate the proportion of the area that will significantly increase in precipitation or runoff in the future. Red and blue represent the results after and before constraints, respectively. The color fill of the basemap represents the difference between after and before constraints.