Amid the age of extreme climate, massive wildfires have become extremely common. In recent years, fires in California, Canada, Oregon, Texas and more have disrupted the lives of millions of residents – and at times left parts of the continent blanketed in a thick smog for days.

Most recently, the horrific fires near Ruidoso, New Mexico, and many other places this spring and summer have led to the question, “what have we learned about forest fires?” in terms of prediction, detection and suppression. The key to fighting wildfires, regardless of their ignition source, is prediction and then early detection. Modern advances in remote sensing and weather prediction make it possible to do all of these with some degree of accuracy.

Prediction

It does not take much of an imagination to figure out that wildfires need dry conditions. Anyone who has ever tried to light a campfire or fireplace with damp wood can attest to this fact. However, there is much more to fire prediction than simply having dry weather.

Any fire, be it a wildfire, campfire or combustion in a gasoline engine, requires fuel, oxygen and an ignition source. While the ignition source can be hard to predict (who knows when a campfire or carelessly discarded cigarette will catch), and oxygen is readily available, predicting the fuel sources is the key component to wildfire prediction.

The purpose of prediction is to warn residents about any impending threats, as well as to help government officials and emergency responders position properly ahead of any wildfire. It also helps them plan for evacuation routes for both residents and escape routes for emergency responders should the fire grow more quickly than expected. A good prediction ultimately saves lives, property and resources.

Conditions for wildfire

Fuel sources vary, but for wildfires, dry brush or grass, dead trees and a few other sources are ideal. One of the first steps in wildfire prediction is understanding which fuels are available. This is often performed through geographic information system (GIS) data that is collected via satellite. Remote sensing via satellite is used to determine where different types of grasses and woodlands are located.

One method for this is by comparing the amount of reflected visible light to the amount of reflected near-infrared light from vegetated areas. Plants may turn yellow when they do not have enough water, but that is only after they have been drastically affected. How infrared light is absorbed or reflected, especially when compared to previous data sets, can indicate a drought before plants actually die. For more information about this, check out NASA’s description on drought prediction.

Once information about the types of fuel is collected, its status can be regularly updated based on weather conditions and biological variables. For example, if an invasive bark beetle has killed a large patch of trees, this will be a particularly dangerous fuel source, as the wood is dead and more likely to be dry.

Meteorological tools

In terms of meteorology, the key factors are the dewpoint (and its impact on relative humidity) and the winds. Strong, dry winds behind dry air masses are the worst scenario for wildfires. This type of data is summarized by the Storm Prediction Center (SPC) in their Fire Weather Outlook, though it is just a starting place for wildfire prediction.

The Storm Prediction Center (SPC) Fire Weather Outlook (6-30-24) shows Elevated and Critical Threat areas, as well as “ISODRYT” meaning isolated, dry thunderstorms that can serve as ignition sources. Source: Storm Prediction Center

From there, the Fosberg Index is a 0 to 100 scale, describing the behavior of wildfires. In this scale, the maximum value of 100 would be attributed to zero moisture content and winds in excess of 30 mph, and would mean that a fire can start or is likely to spread quickly.

Fosberg Index- the ring of 90 corresponds to 23 mph winds and a relative humidity of around 10%. Source: Storm Prediction Center

There are also several Haines Indexes (adjusted for different altitudes) that focus more on the qualitative speed at which an existing fire can grow. These look at the atmospheric stability (temperatures at different levels in the atmosphere) and dewpoint depression (or rather how low the dewpoint is relative to the air temperature).

Worst-case scenario

Perhaps the worst-case scenario is high winds behind a dry air mass. Vegetation has had several days (or weeks) to dry out, providing ample fuel. An upper-level trough is moving into the region. Ahead of the trough, winds will increase significantly, and deep mixing between atmospheric layers will limit the moisture.

Furthermore, there is a little bit of instability, meaning a few dry thunderstorms may form. The dry thunderstorms will be too dry to allow rain to fall. Instead, rain will evaporate upon its descent to Earth. As it evaporates, it cools the surrounding air, causing it to sink. The more evaporation, the more severe the cooling, and the faster the air sinks. Once it hits the ground, it spreads out rapidly as gusty winds that can fan the flames of any existing spark. These dry microbursts are sometimes referred to as a “haboob,” especially if it kicks up a lot of dust, or a “virga bomb,” as evaporating rain is called virga.

The instability also adds the possibility of lightning strikes. Many wildfires are started by lightning strikes, and so a dry lightning strike in a windy, dry-fuel-rich environment is a recipe for disaster.

Prediction during the worst-case scenario

Thankfully, meteorologists are not unaware of these threats. When an area has been dry for some time, this has typically been tracked. The trough pattern, which often puts strong winds behind a dry air mass, can be tracked for many days at a time, and global weather models, such as the Global Forecast System (GFS) model are good at predicting this widespread pattern.

The days with the largest threat can often be predicted using the model data, as well as weather balloon soundings. One of the classic ways to predict dry thunderstorms is with an “inverted V” sounding, where the temperature and dewpoint (red and green traces, respectively) form an upside-down V. This indicates dry, hot air at the surface, but then some area aloft where the humidity is higher — high enough to form a storm, but as the rain falls through dry air it will evaporate.

Inverted-V sounding. Where the red and green lines are close together, the humidity is higher- perhaps a storm could form, but then there is nothing but dry air underneath of it. Source: Storm Prediction Center

Conclusion

Determining how to predict wildfire initiation and spread will always be a challenge. Some of the challenge is figuring out how to quantify the threat, then use those numbers to categorize the appropriate preventive actions. As meteorological data improves, remote sensing techniques expand and higher-resolution models become available, there is some hope that humans can prepare for wildfires as they continue to expand into wild and forested areas.