Supercomputer predicted prolonged drought in United States after 27 years

The climatic model developed by the researchers of the Argon National Laboratory of the US Department of Energy, predicts long-term droughts on Most of the US territory, which will follow short -term but destructive floods.

according to forecasts, all cataclysms will occur over the next 50 years (once every 10, 25 and 50 years). For example, the probability of drought in 50 years is 5%, and it will be widespread in the Midwest, South-West and north-west of the United States. According to scientists, the area of ​​drought increases, and by the middle of the century, after 27 years, a significant part of the Midwest will be in a state of constant drought. The model predicts such a climate in blocks up to 12 square kilometers.

A fragment of the map of the drought areas in the West and Middle West of the United States.

The climatic model was modeled using supercomputers in the National Laboratory of the US Department of Energy in Berkeley (LBNL) and at the Argon computing center of leadership (Alcf).

The work is designed to help lawmakers make reasonable decisions in the development of a policy regarding short -term drought and floods. However, the work has just begun, and now the team is working to improve the resolution of the model up to 4 square kilometers, and uses machine learning methods to identify both short -term and long -term problems.

The group limited the long-term trajectory of these forecasts for 50 years due to a high degree of uncertainty. In predicting the conditions of drought, it is necessary to take into account more than 50 indicators, including temperature, precipitation, evaporatranspiration, etc. To reduce the number of indicators taken into account, the researchers have developed a new measure – a standard index of pressure pressure deficiency (SVDI), which is calculated regardless of precipitation.

According to scientists, many equate a decrease in the amount of precipitation to drought, but this is not always a good indicator. SVDI allows you to measure the need for evaporation. The higher it is, the more moisture is pulled out of vegetation and soil. Another advantage of SVDI is that the pressure deficiency of steam pressure is relatively easy to simulate.

Argonan researchers also predict short but intensive precipitation periods – characteristic of many areas prone to drought – which lead to extensive floods. Also, an increase in water will be caused by drying of the soil – as the soil dries becomes hydrophobic and begins to intensively “push” water.

Ultimately, scientists hope that improved models will help politicians more effectively solve climate problems.

/Media reports cited above.