Seasonal rainfall forecasting - WCSSP case study

The Yangtze River Basin is home to more than 400 million people and is a key agricultural region in China. The region is prone to flooding on a semi-regular basis due to the influence of the East Asian Summer Monsoon. Researchers from the Climate Science for Service Partnership China project have co-developed a seasonal forecasting service for summer rainfall in the Yangtze River Basin. The service is helping users in the region make early decisions to manage flooding in addition to supporting the production of hydroelectric power which is essential for the large cities of eastern China. 

The Yangtze River is the world’s third longest river spanning 3,900 miles. Intense or prolonged periods of rainfall in the region can lead to flooding and significantly impact livelihoods. For example, flooding in 1998 affected over 200 million people with 15 million farmers losing their crops.  

Yangtze River Basin, China

To help manage river levels, the river has hydroelectric dams which can be operated to withhold or release large quantities of water and control the impacts of flooding and drought downstream. The dams also rely on seasonal rainfall to produce electricity for provinces and major cities in eastern and southern China and managers of the dams can regulate electricity supply by controlling the river water level.  

To support these activities, having a forecasting system that provides advance information about how much rainfall there could be over the summer months is vital.  

Developing a prototype service 

Through the CSSP China project, researchers in the UK and China have co-developed a seasonal rainfall forecasting service for the Yangtze River Basin. In late 2015, researchers identified that the Met Office seasonal forecasting system has significant skill in forecasting the variability of monsoon rainfall in the region, which is dominated by the influence of the El Niño Southern Oscillation (ENSO). Strong El Niño events tend to lead to above-average rainfall and flooding in the Yangtze River Basin in the following summer.   

The researchers then worked in collaboration with decision makers to develop a prototype service that delivered forecasts of summer rainfall for the Yangtze River Basin up to a month in advance. The first trial forecast was issued in 2016 and was trialled by collaborators at the China Meteorological Administration (CMA) and water management organisations in the Yangtze Basin, including the Three Gorges Dam operators, whom CMA advise. 

Summer rainfall forecasts can support decision making around the operation of the Three Gorges Dam.

“It was an exciting moment to see one team of scientists identify the user need for an improved forecast service, while another team discovered that our seasonal forecast system had useful levels of skill in the region. The strong El Niño developing at the same time provided an ideal opportunity to develop and test a new forecast service.” Philip Bett, Senior Scientist at the Met Office. 

Since then, the researchers have continually developed the service in response to user needs. Users requested longer lead times and more regional detail, with a need for information in both the upper and middle to lower Yangtze regions to aid hydrological planning. Further scientific research identified how this could be achieved, and by 2019 the project had incorporated this into the service, with forecasts of summer rainfall now provided up to three months in advance.  

In addition to this, in 2020 the service started providing an individual prediction for June mean rainfall, as this period often sees the most rainfall in the region.  

Project impact 

The climate service was found to be useful, with users highlighting how it had helped them develop flood control plans that prevented flooding and avoided agricultural losses.  The researchers are still advancing the service and in 2022 they found that there is good skill for forecasts of the likelihood of enhanced or reduced monsoon rainfall for May–June–July as far as six months in advance. The project team are now looking at trialling this extension to the range of forecasts in the future.