Several storms exhibited highly unusual patterns, forcing meteorologists to exercise extreme caution in tracking, analyzing, and issuing forecasts.

In response to this alarming escalation in natural disasters, VietNamNet sat down with Dr. Hoang Phuc Lam, Deputy Director of the National Center for Hydro-Meteorological Forecasting (under the Hydrometeorology Authority, Ministry of Agriculture and Environment), to assess the 2025 storm season and discuss projections for 2026.

Vietnam witnessed extreme weather events in 2025, especially storms, torrential rains, and floods. Could you provide an overview of this situation?

As of the end of November 2025, the East Sea had recorded 21 tropical systems (15 typhoons and 6 tropical depressions), making it the most active year on record, surpassing the previous high of 20 storms in 2017.

Moreover, in October and November, northern and central Vietnam were struck by multiple widespread heavy rain events. From October 6–8, heavy rainfall triggered by Typhoon No. 11 and its remnants drenched the northeastern provinces, particularly Tuyen Quang, Thai Nguyen, and Hanoi.

Central Vietnam suffered two particularly prolonged and intense rain episodes. The first, from October 23 to November 3, affected provinces from Ha Tinh to Quang Ngai. Notably, Bach Ma station recorded 1,740mm of rainfall on October 27 - the highest daily rainfall ever measured in Vietnam.

The second wave occurred from November 16–21, impacting areas from Quang Tri to Khanh Hoa.

This extreme rainfall led to widespread catastrophic flooding, with water levels on numerous rivers surpassing historical records.

Never before had Vietnam recorded both "exceptional" and "historic" floods on 20 rivers in the same year. The most severe incident was a historic flood in Phu Yen (now part of eastern Dak Lak province), which exceeded the record flood peak of 1993 on the Ba River basin.

In the face of these powerful storms, forecasters held critical discussions. Could you describe a particularly difficult moment in forecasting this year?

One of the most memorable cases was forecasting Typhoon No. 15. On the morning of November 25, while the system was still a tropical depression near the central Philippines, meteorological centers across the region and globally - Vietnam included - predicted it would strengthen into a typhoon upon entering the East Sea.

However, the forecasts diverged significantly in both trajectory and intensity.

At the same time, models were signaling the possibility of a widespread, heavy rain event lasting 3–5 days across south-central Vietnam.

This was a highly sensitive situation, as the region had just suffered devastating floods. Failing to issue timely warnings could leave communities unprepared, potentially multiplying the damage. On the other hand, if warnings were issued prematurely and the storm failed to materialize, it could sow panic among already-vulnerable populations.

This kind of decision placed enormous pressure on the forecasting teams.

As we know, late-season typhoons often interact with cold air masses and form at lower latitudes, making them notoriously difficult to predict.

After extensive collective analysis of rainfall potential, wind intensity, and expected impacts on coastal infrastructure, transportation, and agriculture, we issued a measured forecast: the typhoon was likely to weaken over the sea, pose limited wind danger to land, and generate only moderate rainfall.

Looking ahead to 2026, as climate change continues to drive complex patterns, what does your agency forecast?

In early 2026, the ENSO phenomenon is likely to remain in a weak La Niña phase, possibly returning to neutral conditions in the second half of the year.

During January and February - the heart of winter - northern Vietnam is expected to face strong cold air events, with risks of severe cold spells, frost, and even freezing rain in mountainous regions.

Conversely, the Central Highlands and southern regions may see unseasonal rain even during the dry season.

Tropical storms and depressions in 2026 are expected to occur at levels similar to the long-term average.

Heatwaves are also likely to be in line with typical years. In the North and Central regions, hot weather will concentrate between April and August, peaking in June and July. In the Central Highlands and South, heatwaves are expected from February to May, especially March and April.

Rainy seasons will start at typical times: around early May for the North, South, and Central Highlands; late August for northern Central Vietnam; and late September for the southern Central provinces.

With increasingly unpredictable and dangerous weather patterns - such as intense short-term rainfall, flash floods, and landslides - local authorities and residents must stay updated on meteorological forecasts and warnings to prepare appropriate response plans.

How are emerging technologies helping improve disaster forecasting?

New technologies in meteorology - ranging from modern observation systems, satellites, and radar, to numerical weather prediction models and artificial intelligence - have greatly enhanced the quality of forecasts and warnings.

On the monitoring side, ground-based observation networks and remote sensing via satellite and radar now update data every 5–10 minutes, enabling continuous real-time tracking across the entire country and coastal waters.

In terms of forecasting, high-resolution numerical models running on supercomputers now provide detailed predictions up to 10 days in advance. These tools have significantly improved the ability to anticipate hazardous weather systems such as typhoons, torrential rain, or cold air outbreaks.

For example, the Hydrometeorology Authority has implemented a high-resolution forecasting model for Southeast Asia and the East Sea, running on a CrayXC40 supercomputer. This enables better alerts for potential areas of extreme rainfall and strong winds.

Recently, artificial intelligence has begun to assist in forecasting, though still mainly in a support role. AI can generate storm and rainfall forecast products, automatically integrate real-time observations, and enhance presentation of forecast outputs.

One practical example is typhoon tracking: satellite data from Japan’s Himawari (updated every 10 minutes), combined with radar and numerical models from Europe (ECMWF), the U.S. (GFS), Japan (GSM), and Vietnam (WRF-DA), allow for continuous monitoring and more accurate predictions of storm trajectories and intensities.

AI algorithms help automatically identify storm centers and integrate observation fields into forecast products, speeding up information delivery and supporting early warning efforts.

Bao Anh