Wildfire scientists have unveiled a new rapid risk-assessment tool that shows substantial wildfire danger to homes and buildings in three major Arctic communities, revealing that between 14% and 26% of structures in Anchorage, Fairbanks, and Whitehorse face high or very high risk from approaching flames. The study, led by Jennifer I. Schmidt of the University of Alaska Anchorage together with Robert H. Ziel, Monika P. Calef, Anna Varvak, and Julio C. Postigo, was published in the International Journal of Disaster Risk Science. Rather than relying on complex fire-behavior simulations that demand specialist analysts and expensive software, the approach combines datasets that most local governments already possess: building footprints, zoning maps, and readily available vegetation layers. The result is an updatable, affordable portrait of which buildings face the greatest threat and where mitigation dollars can do the most good.

The urgency behind the work is hard to overstate. Wildfire costs and damages in the United States alone are now estimated at between USD 394 billion and 893 billion annually, and the 2019 to 2020 Australian fire disasters cost nearly USD 100 billion. The Arctic, often imagined as too cold and wet to burn, is not immune: since 1980, wildfires have cost Alaska between USD 2 billion and 5 billion, with projections suggesting costs could rise 68% by the end of the century. Canadian suppression expenses are expected to climb by 60 to 110% relative to the 1980 to 2009 baseline. Meanwhile, structure loss from wildfire has doubled over the past two decades in the western United States, and western Canada has suffered major structure-loss events in four of the last eight fire seasons. Nearly half the global population now lives within the wildland-urban interface, the zone where people and forests meet or intermingle, making tools that translate fire science into actionable local maps increasingly vital.

The researchers’ central premise is conceptually simple but technically consequential: for a structure to be at high risk, a flammable hazard must exist nearby, within roughly 500 meters, and the surrounding landscape must exhibit high wildfire exposure. The method begins with a flammability hazard layer at 30-meter resolution, built from a modified version of NASA’s ABoVE Landcover dataset in which generic evergreen classes were refined to reflect locally flammable species such as black spruce, white spruce, lodgepole pine, and subalpine fir, each assigned hazard ratings from 0 to 100. Critically, the team treated structures themselves as fuel. Each building contributes three fuel units to the surrounding flammability score, a value chosen after systematic testing of one through ten fuel units; three proved optimal, nudging densely built environments up a hazard category without causing wholesale reclassification. Structure density was computed at 90-meter resolution, multiplied by three, and added to the vegetation-based hazard to produce an enhanced flammability hazard layer.

Wildfire exposure, the second pillar of the assessment, captures the potential for hazard to travel. It is calculated as the average flammability hazard within a 100-meter or 500-meter radius, representing the likelihood that fire can reach and burn a given location even if the local fuels themselves are sparse. Exposure values, also ranging from 0 to 100, are classified into five tiers from very low to extreme, and concentrated pockets of flammable vegetation drive scores upward. The third ingredient is structure vulnerability, produced by multiplying the enhanced flammability hazard by a structure importance factor derived from zoning data: residential buildings and critical infrastructure such as hospitals, schools, and nursing homes receive the highest weight, followed by commercial and other structures. Where zoning was unavailable, as on military installations like Joint Base Elmendorf-Richardson in Anchorage and Fort Wainwright and Eielson Air Force Base in Fairbanks, the team collaborated with base personnel or interpreted aerial imagery to classify neighborhoods. The final risk score synthesizes hazard, exposure, and vulnerability into six categories, from extremely low to very high.

The results paint a nuanced picture of Arctic fire danger. Fairbanks showed the greatest proportion of structures in the top two risk categories at 26%, followed by Whitehorse at 22% and Anchorage at 14%. Anchorage, paradoxically, had the highest share of buildings in the top vulnerability tiers, 42%, yet the lowest risk, a consequence of its large urban footprint where dense housing raises vulnerability but fragmented, less-flammable vegetation and abundant developed ground suppress exposure. Fairbanks and Whitehorse, with structures woven into spruce- and pine-dominated boreal landscapes, showed closer alignment between vulnerability and risk. Anchorage was also the only study area where more than 1% of cells jumped two or more flammability categories when buildings were counted as fuel, at 7%, reflecting its 27% share of structure area containing seven or more buildings per cell. The analysis also confirmed the primacy of exposure: when a structure’s wildfire exposure was extreme, at least 87% of such buildings landed in the highest risk category across all three communities.

A key innovation lies in how the team redrew the wildland-urban interface itself. Traditional WUI maps distinguish urban, interface, and intermix zones; the researchers added a fourth category, the transmission zone, marking areas roughly 2.4 kilometers beyond the WUI boundary where fire could move into a community and where fuel breaks might be sited. Within the 500-meter buffer around structures, urban areas were defined by exposure values of 20 or less, interface zones by 21 to 60, and intermix by values above 60, cross-referenced with Alaska and Yukon wildfire management designations that prioritize suppression from critical down to unplanned or wilderness levels. Fairbanks claimed the largest overall WUI footprint, and the transmission zone proved the largest category in every study area, spanning 23,347 square kilometers around Anchorage, 6,180 square kilometers around Fairbanks, and 665 square kilometers around Whitehorse. Across all sites, most structures fell within interface zones, 52 to 61%, followed by intermix at 20 to 35%. Thousands of buildings sit in the high-risk interface and intermix zones where mitigation should concentrate: 15,537 structures in Anchorage, 16,212 in Fairbanks, and 3,734 in Whitehorse.

The method faced its most demanding test against a real disaster. On 17 August 2019, during southcentral Alaska’s hottest and driest summer on record, the likely human-caused McKinley wildfire was driven by high winds across 13.3 square kilometers in three days, destroying 52 residential structures, 84 outbuildings, and three commercial buildings, affecting 1,028 properties and forcing roughly 350 to 400 people to evacuate. By comparing pre-fire aerial imagery from 2017 with post-fire imagery from 2020 at 0.3-meter resolution and linking vanished buildings to the Matanuska-Susitna Borough structure database, the team identified exactly which structures burned. The verdict was statistically compelling: burned structures had significantly higher wildfire exposure than those that survived, and the proportion of burned structures with exposure values above 60 was significantly larger than among unburned ones. Structure risk scores above 60 were also significantly associated with burning. Notably, the test held even under conditions so extreme that normally resistant, low-flammability fuels were igniting.

The study’s design was shaped by sustained community engagement. Through workshops in 2019, 2021, and 2023, along with 15 to 20 meetings organized by Anchorage community councils and landowner associations and public events ranging from wildfire awareness days to Arbor Day, residents and agency staff reshaped everything from color schemes to model parameters. One consequential example: Anchorage residents and fire managers insisted that the maps reflect the spruce beetle outbreak first detected in 2016 and peaking around 2020, which killed and removed many spruce trees. The team responded with rules derived from aerial detection surveys, reclassifying white spruce cells in areas of beetle intensity of 11% or greater to mixed forest, and shifting mixed forest cells to open shrub, while a succession model updated vegetation on areas burned within the last 30 years. This iterative dialogue ensured outputs that both wildfire practitioners and the general public can interpret, a prerequisite the authors see as central to building fire-adapted communities.

The practical implications for planning are direct. Because mitigation funding is often tied to officially designated WUI areas, the ability to rapidly recalculate boundaries from updated building layers expands funding opportunities and keeps pace with community growth. The framework also points to different interventions in different zones: structures in intermix areas, surrounded by vegetation, benefit most from fuel treatments, while those in interface zones, where flammable vegetation is scarcer, are better served by home hardening measures such as fire-resistant roofs, siding, decks, screens, and unburnable buffers. Given that wildfire activity tends to occur where exposure exceeds 60, the authors recommend concentrating resources on the top two risk categories. Where national products like wildfirerisk.org rely on generalized fuel layers that can overestimate unburnable area in WUI zones and are updated too slowly for annually burning landscapes, this approach trades complexity for transparency, using local data that governments already collect. The underlying wildfire exposure toolbox is publicly available on GitHub, and the team argues that empowered residents, equipped with accurate, current, neighborhood-scale information, are crucial to reducing wildfire risk, particularly where communities may prefer action without government intervention.

Affordable, fast-heating and with a clever air fryer function, this versatile model is a winner for outdoor cooks who are willing to give up the traditional flame

• The best pizza ovens – tested
• Gozney Dome Gen 2 pizza oven review

An electric pizza oven with no flame and no need to rotate the pizza may seem sacrilegious to purists, but the Ninja Artisan electric outdoor pizza oven solves many of the hassles of outdoor pizza making. There’s no lighting or monitoring a flame, no need to keep adjusting the heat, and no need for a turning peel, bulky gas bottle or constant use of an infrared (IR) thermometer (although you may still want one – read on for details).

Equipped with four presets – Artisan, Thin Crust, New York, Pan – plus a custom setting to enable you to bake your pizza exactly how you like it, the Ninja Artisan is as effortless as can be, even next to the simplest plug-and-play gas oven. As you might expect from Ninja, it’s a multipurpose cooker, too, transforming into an outdoor air fryer with a capacity to turn out up to 1.3kg of french fries, or bake six chicken breasts.

Stick vacuums are a convenient alternative to corded designs, but which model wins for overall cleaning prowess? Our expert reveals all

• The best robot vacuums
• How to make your vacuum last longer

Choosing a cordless vacuum isn’t a decision that should be taken lightly. You’re likely to keep a vacuum cleaner for years, relying heavily on its ability to suck up dust, crumbs, mud, pet hair and any other dry spillages or sheddings that end up on your floor. Choosing the right model can be the difference between an effective cleaner that’s a delight to pull out of the cupboard and a dud that you dread having to unblock, detangle and clean after every use.

In this review, I took 16 of the leading cordless vacuum cleaners from a range of manufacturers and at various prices and inflicted the same cleaning tests on each one. That takes all the guesswork out of picking your next cleaner: I can tell you exactly which ones picked up the most mess.

Best cordless vacuum cleaner overall:
Shark PowerDetect Clean & Empty IP3251

Shark’s cordless vacuum is a whiz at cleaning hard floors and pet hair, but a disarmingly short battery life caught our reviewer off guard

• The best cordless vacuum cleaners – tested

Vacuuming rarely starts with vacuuming. First you have to clear the floor of stray socks and abandoned toys left behind by children, pets or messy partners. Then there’s the irritation of finding the vacuum’s dustbin still full of dirt, or the battery flat because someone forgot to charge it.

The (take a deep breath – long name coming up) Shark PowerDetect Speed Clean and Empty Pet Pro cordless vacuum Ultramarine IA3241UKT certainly won’t pick up the scattered toys for you, but it has been designed to make the job of vacuuming as frictionless as possible.

Whether you’re on PC or a console, level up your setup with our gaming expert’s pick of the best chairs

• The best games of 2026 so far

Comfort is crucial to the gaming experience. It’s hard to focus on blasting aliens or taking Stoke to the Champions League final if your bum has gone to sleep because you’ve been sitting on an old chair for three hours. We’re spending more time playing games than ever before, so good posture matters. If you have set up a dedicated gaming space, then a rickety seat just won’t cut it.

From luxurious ergonomic thrones to cheap-and-cheerful bucket seats, the options available are bewildering. So I set out on an epic quest: to find the most comfortable, durable and stylish gaming chairs. Here’s what I discovered.

Best gaming chair overall:
Secretlab Titan Evo

Best budget gaming chair:
ThunderX3 Solo 360

Surprisingly easy to manoeuvre and skilled at cleaning carpet and hard floor, Shark’s upright bagless model is one of the best corded vacuums I’ve tested

• The best vacuum cleaners

Corded vacuum cleaners aren’t as glamorous as their cordless cousins, but swapping battery power for a plug socket has its advantages. The device will never be short of power, so it can keep cleaning for as long as you need it to. Prices also tend to be lower because there’s no expensive battery to factor into the cost.

On the downside, there’s a cable to manage while you clean, and even the longest cords usually need to be switched from one plug socket to another if your home is any larger than a small flat.

As summers become hotter, air conditioner sales are booming. If you’re looking to invest, here’s what to consider

When a heatwave struck the UK this week, Jon Connorton, a software developer, began monitoring temperatures inside his east Hampshire terrace house. With some rooms reaching close to 40C, it was time to deploy the air conditioner. “We just wheel it out in emergencies,” he said. “We were having trouble sleeping.”

Connorton and his wife have a portable air conditioner. These plug-in devices cool interior air by removing heat from it and blowing that heat outside, typically via a large hose slung from a window or door.