Projection by international research team from NIES, the University of Tokyo and others sees drought exceeding record levels becoming the norm

How will droughts change in the future due to global warming? Knowing this is an important decision-making tool for considering long-term measures to counter global warming. Knowing when conventional statistics and experience are no longer applicable is vital, especially in the water resources, agriculture and energy sectors. An international research team from the National Institute for Environmental Studies, the University of Tokyo, the Korea Advanced Institute of Science and Technology, and others analyzed global future forecast data on river discharge using a numerical model to investigate the frequency of droughts and estimated for the first time in the world when droughts exceeding current record highs will occur continuously for many years. In other words, they estimated the time when "abnormal" will become the norm. Their study was published in Nature Communications.

The research group defined the "time of first emergence (TFE) of consecutive unprecedented drought" when the frequency of droughts in each area continues to exceed the maximum of the past reference period (1865-2005) for a number of years as "the time when "abnormal becomes the norm." For example, they denote a "period of at least five consecutive years exceeding the maximum historical drought frequency" as TFE5. The paper estimated the TFE for each of the 59 global regions defined based on watersheds, with multiple minimum duration of consecutive exceedances. Their analysis uses hydrologic simulation datasets calculated by a global research group in the Inter-Sectoral Impact Model Intercomparison Project, which analyzes global river flow data from 1861 to 2099.

Here they use five global water model climate projections, each calculated using four climate model climate projections as input data. They examined two global warming scenarios: a scenario for achieving a decarbonized society (RCP2.6) in which all possible global warming measures are applied, and a scenario of advanced global warming (RCP8.5) in which no further global warming measures, such as CO₂ emission reductions are applied, to evaluate differences in outcomes depending on the choice of global warming measures.

Their projections predict statistically significant increases in drought frequency for 25% and 28% of the global land area under the RCP2.6 and RCP8.5 scenarios, respectively, around the middle of this century, with frequency more than doubling in some regions. In both scenarios, the Mediterranean basin, southern and central South America, and Australia would be hotspot areas of increased frequency. These areas have shown a steady increase in drought frequency, which as of mid-century is already significantly greater than in previous reference periods.

In the RCP8.5 scenario, TFE5 was detected in 18 of the 59 regions in this century. Even under the RCP2.6 scenario, which assumes the promotion of global warming mitigation measures, TFE5 was detected in 11 regions during this century. In addition, seven regions (RCP 8.5) and five regions (RCP 2.6) are projected to reach TFE5 within the next 30 years or so. Southwestern South America, Mediterranean Europe and North Africa in particular had noticeably early TFE5 values in both scenarios. On the other hand, the differences between scenarios seen in the TFE5 estimation results indicate that, in general, the choice of the two warming pathways makes a big difference in the timing of TFE5 and its confidence level. In other words, the study also showed that achieving a decarbonized society (RCP 2.6) is important for reducing the probability of or delaying the onset of unprecedented drought conditions.

By comparing time-series changes in future droughts with the maximum values of a reference period, the study successfully estimated for the first time in the world when "abnormal" in the past will become normal due to intensifying droughts and reach a state never experienced before. The research demonstrated that there are large regional differences in the timing, and in some areas the timing was predicted to be on a time scale of about the next 30 to 50 years. In areas where the TFE has been detected, adaptation measures need to be pushed forward efficiently and quickly.