So I recently came across a theory about the dispersal of alumina into our environment that would turn our forests into roman candles and help bring about the firestorms theorized by climate scientists, of course caused by global warming. The theory goes that the alumina, or aluminum oxide, being released into our environment would cause the woody plants that get coated in it and take it up to be turned into a type of thermite. Thermite reactions burn extremely hot, and one type of thermite is even used to weld pieces of steel railroad track together. I looked this up and it turns out that carbon feedstocks actually can be used as a reducing agent in a carbothermal reduction of alumina. It's actually being experimented on right now to be used in place of the energy-intensive Hall-Heroult process for electrolytically reducing alumina (aluminum oxide) into metallic aluminum. The problem they're running into however is the extremely high temperatures required to start the reaction. Unlike other types of thermite, it's difficult to get this one to go at normal temperatures, and for that reason, it hasn't really taken off as a commercially viable alternative to the Hall-Heroult electrolytic process.
Wildfires typically burn at 800 degrees C. The upper range that link claims is 1200 degrees C. Carbothermal reduction of alumina needs to reach 1500 degrees C to start (see link in the above paragraph on carbothermal reduction of alumina), quite a bit hotter than those estimates. To give a frame of reference, a high fire kiln for firing clay tops out around 1300 C, and above those temperatures those clays and ceramics will begin to melt. Lava tops out around 1200 C when it comes out of the ground, and the conditions required to get the fuels found in forest fires even that hot would have to be absolutely perfect, so suffice it to say, it sounds like this one is bunk, since you'd already be looking at the conditions for a firestorm even before you got to the temps required to start this reaction. Short of something like a nuclear detonation, we'll never see those temperatures in our wilderness fires, and if we've got nukes going off, I think carbothermic reduction of aluminum in the burning trees is going to be the least of our problems, considering that nukes can ignite their own firestorms regardless of what fuels are available. Phew! One less thing we have to worry about.
Disclaimer: I am not the world's foremost expert on this subject. This is just me taking 20 minutes to look this up and sharing my findings here for anyone who's interested. If you have something to add or find some error in my assumptions, please share below. I welcome civil discussion.