Early warning

This study's objective is o assess the timeliness of outbreak detection, notification and response against the 7-1-7 targets, and to identify the bottlenecks and enablers at each stage.

A decade-long study of earthquake activity in California has provided the strongest evidence to date that forecasting models can reliably describe patterns of earthquake activity over the following day.

2025 landslide-triggered mega-tsunami in Alaska shows growing risks as glaciers melt. Scientists propose a progressive alert system using weather, glacier changes and seismic activity to warn communities and vessels of potential landslides and tsunamis.

Researchers found two new magma sources under the active volcano, highlighting the critical importance of understanding transitions in volcanic behavior and shallow magma processes, especially when they potentially pose hazards to local communities.

The authors introduce a new framework for subseasonal weather forecasts, MARS+Forecast, which extends MARS yield predictions using ECMWF four-week temperature and precipitation forecasts following the MARS publication date.

The October 2026 GEOGLAM Crop Monitor for Early Warning report provides an overview of global crop conditions with a focus on regions vulnerable to food insecurity.

Using the October 2018 extreme sea level event in the Adriatic Sea as a case study, the authors evaluated two different approaches to forecasting compound sea level extremes.

India is using sub-seasonal forecasts to help farmers anticipate rainfall, drought and pest risks, with examples showing how longer-range weather information supports planting, harvesting and more efficient irrigation.

Between 2010 and 2025 floods in Indonesia claimed more than 1,700 lives and caused an estimated USD 5.8 billion in losses. ADB-backed forecasting systems now warn communities up to nine hours ahead, alongside livelihood support for affected households.

Households absorb much of the cost of disasters: Post-Disaster Needs Assessments from 20 countries in Asia and the Pacific put about 40 per cent of impacts in social and productive sectors. ESCAP sets out three government interventions.

An Indian Institute of Technology Madras (IIT Madras)-led international study has found that atmospheric air pollution could play a critical role in improving the forecasting of rainfall and urban floods.

A roving crew installs cameras on California’s highest, craggiest ridgelines to catch wildfires early. Since 2023 artificial intelligence has scanned the network’s footage, and it now spots 51 per cent of the state’s wildfires first.

The Indonesian government is working with Japan to broadcast earthquake early warnings through mosque loudspeakers, with trials planned in Jakarta and other regions by 2030 before nationwide rollout.

Federal agencies in the United States already use drones to map wildfires and detect hotspots. The Forest Service is now testing ground robots, and researchers are training machine-learning models on drone imagery to predict how fast a fire will spread.

WMO’s Global Seasonal Climate Update forecasts the Niño 3.4 anomaly reaching about +3.7°C in October–December, the highest since observations began, with a near-100 per cent likelihood that El Niño persists through February 2027.

Firefighter and researcher Marc Castellnou sends weather balloons into towering wildfire plumes to measure temperature, wind and humidity, opening a new way to anticipate when a fire turns extreme – and why some keep burning fiercely at night.

Researchers developed advanced deep learning models incorporating attention mechanisms to identify important patterns in rainfall and river flow data, then combined the models’ predictions using a fusion approach to improve flood forecasting accuracy.

When disasters are imminent, timely warnings can be the difference between life and death. However, warnings are only effective when they are received, understood, and acted on.

The Government of Lao PDR, supported by the United Nations, NGOs and CSOs, launch an inclusive risk communication initiative and a video competition on the International Day for Disaster Risk Reduction to improve early warnings.

This publication highlights the progress made in five countries in advancing the EW4All initiative through targeted investments and technical assistance funded by the Government of Sweden via the United Nations Office for Disaster Risk Reduction (UNDRR).

This policy brief presents recommendations on extreme heat early warning. Heat is an escalating global threat, contributing to nearly half a million deaths annually and impacting multiple sectors including health, agriculture, and livelihoods.

El Niño and warming in the tropical Atlantic could help provide advance warning of extreme heat in the Amazon, with some signals appearing as much as seven months earlier, research has found.

To produce predictions, generative AI models must be primed with existing information called training data, making predictions based on patterns observed in previous data sets.

A publication highlights how WMO’s support is helping Bangladesh, Haiti, Liberia, Mozambique and Somalia strengthen observation, forecasting and early warning capacities as part of the Early Warnings for All initiative.

Seventy-three years after the magnitude 9.0 earthquake of 1952 in the same region, Kamchatka has reminded us that it is one of the most efficient tsunami sources in the Pacific.

This case study examines the current status of Nepal’s heat early warning system (HEWS), focusing on risk knowledge, monitoring, warning dissemination, and preparedness

This case study examines the current status of Bangladesh’s heat early warning system, outlining how heat risk is understood, monitored, and communicated, and how warnings prompt actions across institutions.

Research on post-alert messaging finds that after a false alert people need two things: assurance the threat is over, and a plain statement that they are safe. Attribution comes later.

Days after Tropical Storm Gonzalo flooded parts of Sal, Cabo Verde, institutions and stakeholders from across the early warning chain came together in Praia on 29 September 2026 to validate the country’s national Early Warnings for All roadmap.

Researchers at The University of Texas at San Antonio have developed a self-sustaining, artificial intelligence-powered flood warning system designed to spot dangerous water accumulation at the street level.

The recent disaster in Nepal may be a warning of what's to come for the rest of the world. While scientists are getting better at identifying where the danger lies, knowing exactly when the next mountainside will collapse is a much harder problem.

An interview with Sarah Jones, President of DWD, partnership and what EUMETSAT’s new generation satellites will bring and how Germany's national weather service applies EUMETSAT data and services in essential weather forecasts and climate monitoring.

This paper evaluates the effectiveness of the current alerting systems for hot and cold weather used in England to notify the health and social care sector of upcoming extreme weather events.

National Emergency Telecommunication Plans help governments, operators and responders maintain communications before, during and after disasters, strengthening preparedness, early warning, response and recovery while building resilience.

This study examines the anticipatory action (AA) activations implemented before the 2025 Pakistan Floods and documents implementation experiences, operational learning, and lessons from these activations.

The Korea Institute of Civil Engineering and Building Technology (KICT) announced that it will apply its early detection technology for quasi-stationary linear rainbands and sudden localized torrential rainfall to the Urban Flood Forecasting Platform.

An EU project is building a multi-risk early warning system that forecasts who and what a hazard will affect, not just where it will strike.

This article shows that it is vital to focus on who delivers the early warning messages. Messengers commanding high levels of trust need to be actively engaged to ensure information reaches people frequently and widely.

Hurricane agency forecasts warn at the country-scale and not for one site. This paper introduces LiveCyc, a probabilistic forecasting system that takes any agency forecast and works out the wind distribution at the 1-km scale of a single neighborhood.

Bangladesh has achieved a milestone in the country’s two-year journey towards an effective and resilient early warning system: the launch of its Early Warnings for All (EW4All) National Roadmap.

A NASA, UCSD and NOAA collaboration uses machine learning on satellite navigation data to detect sudden atmospheric moisture spikes, flagging likely flash floods for forecasters in as little as fifteen minutes.

China’s AI-powered weather system deployed in Pakistan to support multi-hazard early warning and disaster risk reduction. Experts welcome its potential but stress that forecast accuracy depends on data quality, local validation and preparedness systems.

This brief highlights Save the Children Nigeria’s Community-Led Early Warning Systems (EWS) and Anticipatory Action approach, which helps communities prepare for and respond to predictable hazards before they escalate into crises.