Issue 616: Long-term global cycles of atmospheric variation can be used to help water managers predict droughts more accurately, up to four years ahead, a new study has found.
As rising temperatures, the growing human population and ecological pressures intensify the impacts of droughts, accurate forecasting is becoming increasingly important. Drought forecasts have the potential to inform water management strategies at national and international level, such as those outlined in the EU’s Water Framework Directive.
However, current forecasting methods, which mostly rely on short-term weather predictions, struggle to predict droughts far in advance, with their effectiveness largely limited to one season. Novel methods are needed to support effective responses to water scarcity. One promising alternative uses knowledge of repeating patterns in major weather systems.
To test this approach, researchers applied data from two sources. Firstly, they used records from the British Geological Survey’s National Groundwater Level Archive (NGLA), captured between 1960 and 2023. They measured drought duration and intensity whenever groundwater levels fell below a range of thresholds.
Secondly, they used the North Atlantic Oscillation (NAO) Index to capture differences in pressure patterns in the atmosphere at sea level between subtropical and polar locations in the North Atlantic. These long-range atmospheric systems, known as teleconnections, can drive weather patterns thousands of kilometres away, and are a leading control on European rainfall.
Researchers compared the drought trends observed in the NGLA with underlying eight-year cycles identified in the NAO Index and tested the significance of associations observed between the two. The combined information was used to build a model exploring how the NAO might influence groundwater droughts.
They found that the NAO patterns can be used to predict drought duration and severity at national and regional levels. The predictions could be made, and planned for, up to four years in advance. This study supports previous suggestions that long range teleconnection trends may be a more reliable predictor of future droughts than monthly data or average winter rainfall.
The model’s performance varied across drought duration, showing the greatest accuracy when predicting longer droughts. Its effectiveness also differed across hydrogeological regions, with the best results observed in areas where the water system responds slowly to changes in rainfall, such as in Sandstone aquifers. The findings are likely due to long-term rainfall deficits driving longer droughts, and regional sensitivity to NAO signals. In the UK, the model performed best in the Sandstone and Thames Chiltern Chalk regions, but had lower predictive strength for Limestone or Southern Chalk.
To mitigate the worst effects of drought, public authorities turn to measures such as restricting water abstraction and transferring water between catchment areas based on supply needs. However, such interventions require coordination between national and regional authorities, as well as clear roles and responsibilities where the private sector is involved in water supply, including monitoring drought indicators, and implementing management actions.
The major drought management component of the EU Water Framework Directive includes the obligation to sustainably manage surface and groundwater. Water abstraction permits should respect ecological flows in rivers, and groundwater abstraction should not exceed inflows. Member States are encouraged to adopt drought management plans. These are not obligatory, but they represent good practice for river basins with water scarcity. These plans promote sustainable water use, including frameworks for water allocation and early warning systems detecting drought risks. More accurate forecasting such as the approach explored in this study could prove particularly valuable.
O artigo foi publicado originalmente em Comissão Europeia.
