Forecasting summer temperature across the Northern Hemisphere: Assessing the predictive limits To what extent can summer surface air temperature over Northern Hemisphere continents be predicted?
//

Forecasting summer temperature across the Northern Hemisphere: Assessing the predictive limits

To what extent can summer surface air temperature over Northern Hemisphere continents be predicted?

Changes in summer surface air temperature have far-reaching consequences for human societies, particularly in relation to the occurrence and intensity of heat waves. These extreme temperatures are directly linked to heat-related illnesses, including heat exhaustion and heatstroke, posing a severe threat to public health. Historical events such as the European heatwaves in 2003 and the Russian heatwaves in 2010 stand as stark reminders of the deadly implications of rising temperatures, causing tens of thousands of casualties. Beyond the immediate health risks, summer heatwave disrupts energy management, agriculture planning, hydrological cycles, and water resources.

Recognizing the critical importance of understanding the drivers behind changes in summertime surface air temperature, accurate prediction becomes imperative for informed decision-making across various sectors. This, in turn, can pave the way for enhanced resource management, improved public health outcomes, and the development of more resilient and sustainable communities.

How predictable are the heatwaves in Northern Hemisphere continents? By identifying and predicting dominant atmospheric circulation patterns through our physically based empirical model, we offer a valuable tool for achieving more accurate seasonal forecasts.

Chunzai Wang

Atmospheric circulations can affect surface air temperatures

Studies have revealed that changes in surface air temperatures during summer over North Hemisphere continents can be strongly affected by the so-called “atmospheric bridge” – a teleconnection between different regions of the Earth’s atmosphere through motions of the air, transmitting signals from one part of the world to another. These atmospheric bridges play a crucial role in shaping weather events, such as droughts, floods, or changes in temperature, in locations far away from where the initial changes occurred. One common type of atmospheric bridge is the atmospheric teleconnection, which involves the coupling of atmospheric circulation patterns over widely separated areas.

There are many atmospheric circulation patterns identified in the Earth’s climate system. Each pattern has different influences on the summer surface air temperature in certain regions. For instance, the summer surface air temperature over Eurasian continents can be affected by the atmospheric circulation patterns over British–Baikal Corridor and Silk Road regions. The Eastern Atlantic, the Scandinavian, and the Western Mediterranean Oscillation patterns can exert notable influences on surface air temperature variations in northeastern Spain.

Our study identified five dominant atmospheric circulation patterns that have important contributions to summer surface air temperature variability over Northern Hemisphere continents. These patterns are all located over the Northern Hemisphere mid-high latitude with distinct characteristics (Figure 1). The summertime surface air temperature over Northern Hemisphere continents can be reproduced well by combining the five atmospheric circulation patterns. Hence, to achieve accurate prediction of changes in surface air temperature, one needs to make forecasts of these atmospheric circulation patterns.

Figure 1. a)-e) The correlation map of the anomalous June-August mean 500hPa geopotential height (color shading) and surface air temperature (contours) with reference of 5 atmospheric circulation patterns respectively.
Credit. Author

Prediction of the atmospheric circulation patterns

Climate models are a powerful tool to simulate and predict the anomalous climate patterns. However, even the state-of-the-art climate dynamical models still struggle to capture the year-to-year variations in climate circulation patterns over the high-latitude Northern Hemisphere. This is because the prediction skill of climate models heavily relies on tropical air-sea interactions. At the same time, the mid-latitude circulation patterns are also strongly modulated by atmospheric internal stochastic perturbations and, thus are much less predictable. 

Considering that the dynamical models may not predict the patterns satisfactorily, the study develops a physically based empirical model to make the prediction. The physical-empirical model has been widely applied for climate predictions, including predicting summer rainfall in Asia, wintertime extreme cold days over East Asia and other climate systems. These studies show that this model is a powerful approach for seasonal prediction compared with the dynamical models, especially for the subtropical to high-latitude land regions.

The first step of establishing a physical-empirical model is to explore the physical processes and predictors that can reproduce the anomalous characteristics associated with the prediction. Unlike pure statistical prediction models, the model in the study aims to identify the most influential predictors based on understanding the physical basis for the lead–lag relationships between predictors and the predictand. To objectively select physically meaningful predictors, we examine three fields, i.e., surface air temperature, sea level pressure, and 500-hPa geopotential height. The first two variables depict anomalous lower boundary thermodynamic conditions, and the last one represents the atmospheric circulation patterns.

Since we aim to predict the circulation patterns during boreal summer (July–August), we search for anomalous signals during April and May to detect predictors in the preceding season. The selection procedure is objective and reproducible, and meanwhile, it is easy to apply to other climate prediction problems. When choosing predictors, we prioritize understanding the underlying physical processes that account for the leading relationships between these predictors and summer atmospheric circulation patterns.

Then, we use the selected predictors to establish the prediction model and to obtain the forecasted circulation indices. The correlation coefficient between our predictions and observations, which evaluates the prediction skill by measuring the similarity between the two, is ~0.5–0.6, suggesting that the five indices can be predicted reasonably well using our model.

Prediction of northern hemisphere surface air temperature

We proceed to make a forecast of the summer surface air temperature using the predicted circulation patterns while fully considering the continued rise of global temperature due to human-induced greenhouse gas emissions. Our prediction results show high prediction skills over almost the entire land areas in the Northern Hemisphere, especially over the western coastal areas and middle North America, north and South East Asia, Europe, Middle East, and Africa, where the prediction skill is rather limited using dynamical models (Figure. 2a).

The predictive power of our model for each year’s surface air temperature change is also estimated. We found that although the predicted patterns of surface air temperature are generally close to observations, the prediction is less successful in certain years such as 1980, 1991, and 2018 (Figure. 2b). The low skills in these years may come from the deficiency of our models or the poor predictability of the atmospheric circulation patterns. A case-by-case study is needed to address the issue further. 

Figure 2. a) The temporal prediction skill for summer surface air temperature prediction over North Hemisphere land areas. b) The pattern prediction skill for summer surface air temperature prediction over North Hemisphere land areas as a function of forecast year.
Credit. Author

Contributions and practical implications of this research

By identifying and predicting dominant atmospheric circulation patterns through our physically based empirical model, we offer a valuable tool for achieving more accurate seasonal forecasts. While acknowledging the challenges posed by certain years, our model demonstrates high predictive skills, particularly in regions where traditional dynamical models fall short. This research contributes to informed decision-making across various sectors, fostering resource management, improving public health outcomes, and promoting resilient and sustainable communities in the face of evolving climatic conditions.

Given the profound impacts of heatwaves on human health, energy demand, and agriculture, it is important to understand the causes and predictability of summer surface air temperature. Based on our findings, we can utilize the prediction model established in this study to make forecasts of surface air temperature during June–August by the end of May each year. The results may lead to better preparedness of the society against heatwave-related disasters. In the future, longer-lead prediction may also be achieved by further considering the influence of tropical oceans, in combination with the predictors based on atmospheric circulation patterns identified in this study.

🔬🧫🧪🔍🤓👩‍🔬🦠🔭📚

Journal reference

Xing, W., Wang, C., Zhang, L., & Zheng, J. (2024). Prediction of summer surface air temperature over Northern Hemisphere continents by a physically based empirical model. Climate Dynamics, 1-15. https://doi.org/10.1007/s00382-023-07065-2

Dr. Wang is an international expert in physical oceanography and climate science. Over his nearly 30-year academic career, he has served as an Oceanographer at the National Oceanic and Atmospheric Administration and a specially appointed researcher at the Chinese Academy of Sciences. His research focuses on ocean-atmosphere interactions and climate change. He has been selected as a highly cited scientist by Elsevier, one of the "Top 2% Scientists Globally" by Stanford University in the United States, and one of the 1000 most influential scientists in the field of climate change by Reuters in the United Kingdom.

Dr. Xing is a research scientist at the State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences. Her research mainly focuses on "Asian monsoon dynamics", including "interannual variability of the Asian monsoon onset", "dynamical processes that influence East Asian monsoon rainfall", and "predictability of Asian summer monsoon rainfall".

Dr. Zhang is a research scientist at the State Key Laboratory of Tropical Oceanography, South China Sea Institute of Oceanology, Chinese Academy of Sciences. He primarily studies "large-scale atmosphere-ocean interactions", with a special focus on "interbasin interaction processes and their climatic impacts". Dr. Zhang has published over 50 papers in journals including Science Advances and Nature Communications. He serves as an Associate Editor of "Geophysical Research Letters" and "Journal of Climate", and is a member of the CLIVAR Climate Dynamics Panel.

Dr. Zheng is an associate professor at the South China Sea Institute of Oceanology (SCSIO), Chinese Academy of Sciences (CAS). His research work focuses on air-sea interactions and extreme climate events. One of his papers has been selected as an ESI Highly Cited Paper. Before joining SCSIO, he completed his PhD in climate science at the Institute of Atmospheric Physics, CAS, and conducted postdoctoral research at the Second Institute of Oceanography, Ministry of Natural Resources.