Since the outbreak of the global COVID-19 epidemic, Collaborative Innovation Center of Western Ecological Safety (CIWES) has actively coordinated various scientific research strengths to successfully develop the Global COVID-19 Prediction System (All copyright by Lanzhou University) on the basis of the existing regional epidemic prediction model. This system has taken the impact of climate and environmental conditions as well as government control measures on the spread of the outbreak into account. Since the launch of the system, our center has been supported and affirmed by the superior units and associate scientific research institutions, and also received extensive attention from the media and the society at home and abroad. As scientific researchers, we have the responsibility and obligation to respond to the concerns of the public and media reporters, and explain to the society the prediction system released by our center. For further information please see the Frequently Asked Questions below.
1. Can you briefly introduce this global COVID-19 pandemic prediction system?
This prediction system is built by combining advanced technology of statistical-dynamic climate prediction with epidemic models, and the impact of temperature, humidity and control measures on the development of this pandemic is considered in the system. NO2 is an important atmospheric pollutant, and its concentration change can quantitatively reflect the strength of government control measures. When the local NO2 concentration is abnormally low, it means that the traffic flow is significantly reduced, and government control measures have played a role in increasing social distance, which will reduce the spread of the virus. Our previous research results show that the optimal temperature for the spread of the COVID-19 is 5-15 ℃, and strict quarantine measures can help reduce the probability of infection. Therefore, three variables, temperature and humidity and government control measures, were introduced into the system to improve the traditional epidemiological model and integrate the achievements of different disciplines.
The goal of the global COVID-19 prediction system is to use scientific methods to predict the development of the pandemic, and provide a scientific basis for assessing the situation of the epidemic and taking effective control measures. We sincerely hope that the system will help the WHO and governments in their response to the pandemic.
2. How to ensure the accuracy of prediction results?
The data in the model construction such as the number of confirmed cases and deaths are based on public data from the Center for Systems Science and Engineering (CSSE) of Johns Hopkins University. Meteorological data such as temperature and humidity come from the European Centre for Medium-Range Weather Forecasts (ECMWF) and National Aeronautics and Space Administration (NASA).The research team combined the advanced technology of statistical-dynamic climate prediction with the epidemiological models and introduced the latest global pandemic data in real time. The system models over 180 countries based on repeated optimal parameter inversion.
Prediction is a scientific issue, which needs scientific, rational, open and cooperative attitude. The accuracy of the prediction results is affected by various factors, such as: the accuracy and authenticity of the epidemic data reported by various countries, the variation of the virus during the transmission process, and the changes of local meteorological conditions (temperature and humidity, etc.) with the atmospheric circulation. At the same time, the control measures adopted by the government are also dynamically adjusted according to the actual situation of each country. Citizens will also adjust their self-protection measures according to the real-time updated epidemic data. These complex factors will affect the development of the epidemic. Therefore, our models introduce the latest epidemic data in real time and constantly update and optimize to ensure the reference value of the predicted results. Currently, we update the prediction results every 10 days.
3. According to the introduction, this system was developed by the Collaborative Innovation Center of Western Ecological Safety (CIWES), which is not a medical research institution. So did you why carry out relevant research on COVID-19?
In the face of the epidemic, we must unite multiple forces, work together to meet the challenges, and be prepared to coexist with the virus for a long time. The monitoring, early warning, vaccine development, emergency response, and resource allocation of the epidemic all require the joint efforts of multiple disciplines. Cross-disciplinary research and cooperation is the right choice in line with the common interests of all mankind. Led by Lanzhou University, CIWES is collaboratively established by related Institutes of Chinese Academy of Sciences, Tibet University, Qinghai University, Gansu Institute of Desertification Control and other institutes. Our mission is to focus on water, soil, ecology, atmosphere and human, ensure ecological security, build a livable environment and protect human health.
Since the outbreak of the global COVID-19 epidemic, CIWES has actively coordinated various scientific research strengths. Under the leadership of leaders of LZU and the center director, we apply our technical expertise and strengths in statistical-dynamic methods for climate prediction to successfully develop the Global COVID-19 Prediction System (All Rights Reserved by Lanzhou University) on the basis of the existing regional epidemic prediction model. This system has taken the impact of climate and environmental conditions as well as control measures on the spread of the outbreak into account. This system was fully supported and coordinated by the School of Public Health, the First Hospital, the School of Basic Medical Sciences and School of Politics and International Relations at Lanzhou University, which is a product of the interdisciplinary integration and cooperation of Lanzhou University. As researchers, we have the social responsibility to use our professional knowledge to carry out research to help effectively curb the development of the COVID-19 pandemic.
4. How to interpret the prediction data and its time-effectiveness?
At present, the prediction results are updated every 10 days, and the prediction can be made for up to a month. In the later period, the optimization will be continued, so as to carry out longer-term prediction, such as seasonal prediction.
The predictions produced by our model are for reference only. It should be emphasized that prediction is a scientific issue, and we must treat it with a scientific attitude. The variables in the prediction model are constantly changing; we will modify the prediction model in real time according to the actual situation in subsequent research. COVID-19 is the common enemy of all mankind. Behind every number are fresh life and happy family. Prediction is more important for early warning. It is expected that governments will adopt scientific and active epidemic prevention policies to contain the epidemic. We do not wish to see the predicted data become reality in any region, and we call on the international community to strengthen cooperation at all levels to jointly fight against the global spread of COVID-19.
5. Differences between seasonal prediction and monthly prediction?
Seasonal prediction uses a more complex and accurate 6-parameter model. It takes into account not only seasonal changes in meteorological factors such as temperature, but also the impact of second outbreak caused by large-scale gathering activities such as protests and city reopening. Seasonal prediction mainly reflects the long-term trend of the seasonal changes of COVID-19, and the prediction time limit is 6 months. The monthly prediction introduces high-frequency components predictions on the basis of trend predictions, and comprehensively considers the impact of changes in the city's lockdown measures and the self-quarantine of citizens on the development of COVID-19, so as to make a more accurate daily prediction. The monthly prediction is updated every 10 days.