Data and information management

This report presents an evidence-based assessment of New Zealand’s atmosphere and climate, highlighting record-high greenhouse gases and temperatures, increased weather extremes, and the urgent need for adaptation and resilience.

Compounding risks to global food system—ongoing conflicts, the most severe El Niño in decades, global trade disruptions—humanitarian data are disappearing. Budget cuts, shrinking development assistance, and competing priorities erode key streams of data.

This report assesses feasibility of long-range flood outlooks for Australia using co-design with end users and BJP model to quantify and map flood likelihood, finding S2S forecasts enhance preparedness by identifying elevated risk weeks to months ahead.

As climate change is expected to increase the risks associated with extreme rainfall and flooding, managing floods requires approaches beyond conventional infrastructure such as dams and reservoirs.

Beyond coordination, a national database could help Australia more fairly distribute bushfire management resources and facilitate the sharing of expertise, equipment and systems.

South Africa launched its first national Climate and Health Surveillance platform, a pioneering research and decision-support tool designed to strengthen the country’s ability to understand and respond to the growing health impacts of climate change.

This report examines the conditions, decisions, actions, and systems that influenced Southern California Fires and community conflagration outcomes before and during the fires.

Insurwave announced a powerful update and relaunch of its dynamic risk platform, bringing insurers, brokers and fleet owners onto one platform designed to make dynamic risk measurable as it moves, alongside a new brand identity.

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.

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.

Team studied tsunami-like sea disturbances in Greece from antiquity to now, finding nine of ten cases were meteorological in origin and only one seismic, stressing the importance of multiple discriminants to distinguish meteotsunamis from seismic tsunamis

This review explores data-driven methods in climate multi-hazard and risk assessment, focusing on four themes: data processing and collection; hazard identification, prediction and analysis; risk assessment; and future risk scenarios under climate change.

The study integrates asset inventories, satellite imagery and impact modelling to estimate flood losses and damages within days, once a sufficiently detailed flood footprint is available.

River models have long been used to estimate water in hard-to-reach or remote areas. New study from UMass Amherst shows the math can be off—way off.

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.

Study analyzes global flood research from 1980–2025, finding surging output since 2000 but persistent mismatches between research attention and actual flood risk, especially in Africa and South America, driven by socioeconomic factors, not risk exposure.

Along Colombia’s San Juan River, Afro Colombian community councils documented earthquake losses house by house – but only municipal authorities can enter those households into the national register that unlocks state assistance.

AI forecasting is quick and cheap, but errors can’t be tracked and explained, and more data and testing are needed

The challenge for AI models to predict hurricane intensity is not just about obtaining more data, building better neural networks or deploying faster computers. It is also about understanding hurricane behaviors and how chaos in intensity emerges.

A framework providing a universal taxonomy for natural hazard and climate risk and resilience assessments of infrastructure systems, defining Risk Classes, Resilience Classes and guidance for hazard, climate and infrastructure assessments.

Coastal flooding seems like an unpredictable event, a natural disaster that occurs randomly. But UCF researcher Ivan Haigh has found that recurrent coastal floods actually occur at highly predictable times.

A new Harvard dataset tracks corporate climate commitments and emissions across 25 years, as US federal reporting programmes are wound down.

Pacific islands need expanded LiDAR mapping of land and seabeds to close critical coastal data gaps, improve sea level rise adaptation, and support better flood modelling, planning and protection of vulnerable communities.

Study reveals a hidden problem: Temperature proxies used in many climate-impact studies can contain substantial measurement error, leading to misestimation of true impacts of temperature and, potentially understate climate impacts by at least 15%.

The DisasterAWARE ecosystem was not introduced in the middle of a crisis—it was already part of the international operating environment that bridges communities. It has been exercised through continuous engagements.

This guidance note provides a framework for evaluating whether climate and disaster risk models are fit for climate and disaster risk financing decisions, covering model transparency, hazard, exposure, vulnerability, loss and uncertainty assessment.

Two smoke-damage bills awaiting the governor’s signature address a longstanding issue brought to the fore by the Los Angeles County fires: When is it safe for wildfire survivors to move back into their homes?

This study develops a standardised, reproducible, and operational methodology to screen tsunami risk at the community scale across the Caribbean Basin, based exclusively on open and globally consistent datasets

More than half of reporting companies provide no core quantitative metrics, raising concerns over their ability to measure nature-related risks and potential greenwashing.

This study presents a detailed and refined glacial lake outburst floods (GLOFs) risk assessment using a data-driven machine learning framework.

This paper introduces the FLOSEV flood severity scale, an impact-based classification system using fatalities, affected municipalities and damage indicators. Tested on more than a century of Italian floods, it enables comparison of flood impacts.

This study examined the sensitivity of flood-frequency estimates to the length of hydrometric records using the annual maximum series (AMS) and partial value series (PVS) approaches at 15 hydrometric stations in the Karkheh Basin, Iran.

This study uses observations and machine learning to estimate bankfull river discharge globally. Results show that bankfull return periods vary significantly across climate zones, challenging the common 2-year assumption used in global flood models.

This paper presents LiveCyc, a machine learning system that converts tropical cyclone forecasts into probabilistic 1-km wind forecasts. Evaluation across 16 U.S. hurricane landfalls shows well-calibrated results and supports optimized warning triggers.

This paper analyses 150 years of census and hazard data to assess how extreme weather affected population trends in Northern Australia. It finds floods and cyclones can disrupt population growth highlighting need for resilience and adaptation planning.

This toolkit examines how extreme weather events fuel mis- and disinformation and provides frameworks for assessing readiness, triaging threats, building partnerships and strengthening information resilience before, during and after disasters.

A deadly Nepal-Tibet border flood highlights how limited cross-border data sharing, restricted access to Tibet and weak early warning systems can hinder disaster preparedness. The article calls for greater regional cooperation on climate and flood risks.

XAIDA is designing and applying AI-based methodologies to reveal how climate change impacts extreme weather events and to foresee their occurrence in the future.