What Are Some Predictions for Future Climate Change?

Climate prediction is based on the evolution of past climate and inferring possible trends in climate development in a certain period in the future. In order to reduce the forecast error caused by the initial field error and imperfect models, the current climate prediction is a ensemble forecast method with multiple initial values and multiple models. Therefore, in essence, the climate forecast is a probabilistic forecast. Climate predictions can be divided into two categories, one using statistical methods and the other by dynamic numerical prediction.

Climate prediction is to infer the possible trends of climate development in a future period based on the evolution of past climate. Because the climate has various time scale changes, from forecasting short-term climate changes within a few decades to forecasting climatic changes over 10,000-year glacial and interglacial periods, all belong to the category of climate prediction. The advent of the glacial period not only changed the entire climate system, but even affected the entire geographical environment. Short-term climate change occurred without changing the geographical environment. The former belongs to climatological time scale climate prediction, and the latter belongs to climatological time scale climate prediction. with

Factors to consider in climate prediction include

Climate predictions can be divided into two categories, one using statistical methods and the other by dynamic numerical prediction.

About 100 years ago, some countries have begun to use long-term forecasts using correlation regression. At present, about 30 countries formally make monthly and seasonal climate forecasts worldwide. In addition to China, the United States, Japan, the former Soviet Union and other countries have more jobs. However, the general level of empirical forecasting is not high. With a relatively rough grading check, the forecast accuracy is only about 55% to 60%. If the inspection is carried out strictly on a monthly basis, this level may not even be reached. Among them, the temperature forecast level is slightly higher, and the accuracy of the precipitation forecast is sometimes less than 55%. ^[1]

1. The latter is a forecast of specific weather conditions, while the former is a prediction of average statistics of climate elements and weather conditions within a certain period of time.

2. Weather forecast is generally limited to the analysis of the physical processes of the atmosphere and hydrosphere, and climate prediction must consider the effects of energy and material exchange and astronomical factors in the climate system including the atmosphere, oceans, continents, ice and snow, and the biosphere.

3. The latter mainly depends on the initial values, while the former depends on both the initial conditions and the boundary conditions or completely on the boundary conditions. The predictability that depends on the above two conditions is called the first type of predictability by Lorenz. For long-term (decades or hundreds of years) climate change predictions, such as global climate change caused by an increase in greenhouse gases caused by human activities, will not depend on the initial conditions of the atmosphere. This is because after long-term integration, the model will Total loss of memory of initial conditions. This type of climate prediction, which is entirely dependent on the changes in detailed boundary conditions, is referred to by Lorentz as the second type of predictability, and its predictability depends on the time scale of external forced change. Due to the inertia of the climate system, the climate system will continue to change for a long time, even as long as the external forcing imposed on the border disappears, even as long as more than ten years. The response of sea level rise is an example. Because the time of the first type of predictability generally does not exceed 3 weeks, short-term climate predictions above the monthly time scale are also basically made in the sense of the second type of predictability. ^[2]

In order to reduce the forecast error caused by the initial field error and the imperfect model, the current climate prediction is a ensemble forecast method with multiple initial values and multiple models, so the climate forecast is actually a probability forecast. ^[2]

On the morning of September 24, 2012, Zheng Guoguang, the director of the China Meteorological Administration, went to the National Climate Center to listen to the climate trend forecasting consultation and gave guidance on the focus of future climate trends. ^[3]

Zheng Guoguang and experts from the National Climate Center discussed the current impact of the development of the El Niño event and the reduction of Arctic sea ice on China's climate. He focused on the fire hazards in the forest areas in the southwest and other regions and climate events that may cause disasters. Zheng Guoguang also provided guidance on how to further incorporate forecast uncertainty into climate services. ^[3]

Experts from the National Climate Center analyzed the recent climate characteristics and the causes of climate anomalies, and introduced the national climate trend outlook for October. ^[3]