Science has good and bad news for the future of wildfires

Image from NASA
By  | 

ANCHORAGE (KTUU) - Scientists have been working feverishly to understand the processes that drive wildfires in the state and in the country, and unsurprisingly there is some worrying news about chance of big fires in the future.

While national agencies offer predictions the data is pretty coarse without nuance to the terrain or ecology of an area.

So researchers like Peter Bieniek of UAF are working to produce better models to increase the accuracy of forecasts and models, particularly on a large scale and in longer time frames. Beniek’s work focuses on analyzing data for future, high-probability fire days, and the results aren’t much of a surprise considering a warming planet.

“Likely looking down the road, especially in these future projections, data show that we're gonna get more higher fire danger over the next hundred years,” says Bieniek.


Photo courtesy of Peter Bieniek.

One of the main drivers of higher probability is a likely increase in lightning strikes in Interior Alaska, which is caused by a higher likelihood of convective precipitation. Convective precipitation occurs when water vapor rises straight up through the atmosphere to form clouds, instead of moving diagonally like normal weather patterns that bring various forms of precipitation.

That moves the total likelihood of a fire year like 2015 -- when five million acres burned -- higher by 34 to 60%, according to Bieniek’s data.

Bieniek’s data show that these sorts of strikes are likely to increase as convection precipitation increases. That, in turn, drives up the likelihood of fires.

An interesting finding is that two different models used by Bieniek and his colleagues indicate that while fire danger is likely to rise in the early season--May, June, and July-- the models diverge as to whether the fire danger will stay as high in August and September. That means that fire managers could be seeing intense strain on firefighting crews earlier in the season, and have it taper off later in the season. For now, though, they aren’t making any bets.

While Bieniek and his research are enough to make the most sunny-eyed managers concerned, other advances in technology and data are unlocking the potential for remarkably accurate predictions of exactly how fires will burn, which could allow managers to avoid confusion about when and where fires could expand.

A National Science Foundation-funded project, NEON (National Ecological Observatory Network), is working on using LiDAR (Light Identification Detection and Ranging) technology to get hyper-accurate maps of vegetation layers that determine how fires will burn.

NEON image of how LiDAR scans can show variation in ground cover

LiDAR scans allow several advantages over traditional methods of surveying and mapping land. Instead of a single pulse of energy returning, as happens in sonar, LiDAR collects information from pulse returns at various wavelengths, which can provide a more nuanced and layered image of a plot of land.

"LiDAR gives us a three dimensional image of the ground,” explained Tristan Goulden, a NEON researcher.

Aerial surveyors fly over plots with instruments including discrete and waveform LiDAR systems, an imaging spectrometer and high-resolution digital cameras, which can provide more information on the exact type of vegetation and its photosynthetic activity. In other words, exactly the flammable birch trees end the slow-burning alders begin could be determined precisely from the air.

By mapping different areas of land with LiDAR, managers one day may be able to tell how fast and where fires might spread. They may also be able to make models with a “combustion completion” ratio that can indicate how much fire is left to burn, allowing managers to make accurate predictions of how long a fire has left to burn, and thus more efficiently allocate resources.

Instead of relying on the relatively course projections derived from thermal imagery, LiDAR combined with hyperspectral imagery, which provides information about the chemical composition of the vegetation, managers could predict where a fire might move. With the recent unpredictability of the fire growth in Southcentral Alaska, better predictions could be crucial for making management decisions.

While Alaska’s high-fire areas haven’t yet been mapped, that could come someday soon. NEON has 8 observational field sites in Alaska, including sites at Toolik Lake on the North Slope, Delta Junction, and Caribou Creek near Fairbanks, which could give officials the tools so that even while the number of fires increases, so too will the effectiveness in fighting them.

And there is also an even more exciting opportunity: NASA just launched two new LiDAR satellites that could help bring virtually limitless mapping to the world’s ecosystems.

Copyright 2019 KTUU. All rights reserved.