The DHCDAI project: Using artificial intelligence to timely forecast droughts and heatwaves in Vietnam

The German-Vietnamese research project DHCDAI is developing new forecasting methods to detect combined drought and heat events several weeks in advance and to enable early countermeasures.

Fig.1: Operational weather station in Hanoi.
In densely built-up urban areas of Vietnam in particular,
the heat stress can be further exacerbated by the urban heat island effect

Droughts and heat waves are among the extreme weather events that are becoming increasingly significant as a result of climate change. The situation becomes particularly critical when both phenomena occur simultaneously and reinforce each other: High temperatures accelerate soil drying, while the lack of soil moisture can in turn intensify the heat. Such combined drought-heat events can have serious consequences for agriculture, water supply, regional ecosystems, and human health. In Vietnam, both densely populated cities - where urban development further exacerbates the heat stress (Fig. 1) - and agriculturally dominated regions are affected. As in many countries around the world, record-breaking heat waves have occurred repeatedly in Vietnam in recent years. In May 2023, the four-year-old record of 43.4°C was surpassed with a temperature of 44.2°C (Fig. 2). By contrast, a drought in the Mekong Delta during the strong El Niño of 2015–2016 illustrates just how severe the effects of drought can be.

So-called “flash droughts,” which can develop within a few weeks or even days, pose a particular challenge. Conventional weather forecasts typically extend only about two weeks into the future, while seasonal climate forecasts provide insights over longer periods. In between lies the so-called sub-seasonal timescale in which forecasts for the coming weeks are generated. This timeframe offers great potential, particularly for the early detection of droughts and heat waves, but has so far been scarcely utilized for operational weather forecasting in Vietnam.

This is where the three-year German-Vietnamese research project DHCDAI (“Development of a Forecasting and Early Warning System for Drought and Heat Wave Hazards in Major Cities and Agricultural Subregions of Vietnam Using Innovative Subseasonal Forecasting Methods”) comes in. In collaboration with the Hanoi University of Science (HUS), the Institute for Meteorology and Climate Research - Troposphere (IMKTRO) aims to develop an innovative forecasting and early warning system for combined drought and heat hazards. The focus is on the metropolitan study regions of Hanoi, Ho Chi Minh City, and Buôn Ma Thuột, while comparative studies are also being conducted for the Rhine Valley in Germany.

The IMKTRO investigates the atmospheric and oceanic factors influencing extreme drought and heat events and evaluates the forecast quality of ensemble forecasts from the European Center for Medium-Range Weather Forecasts (ECMWF). Using modern artificial intelligence methods, the project aims to correct systematic forecast errors and improve forecast quality. The forecasts for the next six weeks are then combined with high-resolution regional model forecasts from the Vietnamese partner, which cover a period of up to two months.

The jointly developed system is initially to be established at the Northern Delta and Midland Regional Hydro-Meteorological Center (NDMRC) of the Vietnamese Meteorological Service and integrated into its operational forecasting activities. Training sessions and joint scientific activities will support staff at the Vietnamese Meteorological Service in applying the new forecasting routine. In the long term, the improved forecasts are expected to help identify risks to agriculture, water management, and the population at an earlier stage and to initiate appropriate preventive measures. At the same time, DHCDAI strengthens the long-standing scientific collaboration between IMKTRO and its Vietnamese partners.

The DHCDAI project started on 01 November 2026.