Letting the fires burns clears out the fuel. IF we go in next year and light that area again there won't be much fuel, so the fire will burn much colder - humans and wildlife caught in that second fire can easily leap over the flames to relative safety (don't get me wrong, some will die in the attempt, but it isn't a death sentence like the larger fires we have now).
That second fire has an additional benefit: carbon sequestration. The big fires we have now turn all carbon to CO2. The smaller fires we could have won't burn hot enough for that and so unburnt carbon will fall to the ground.
The best first-order defenses I’ve read about would be strict code for setbacks around houses to give a decent buffer between homes and forest, possibly some material requirements for roofs and siding, and generally making homes and communities more burn-proof. This obviously doesn’t help if there’s a raging inferno next door, but it shifts the curve.
Second, is better warning and alerting systems, and better evacuation management systems to get people prepared to leave, aware of when they absolutely must leave, and able to leave quickly in the moment. These systems are severely lacking for any natural disaster response, not just forest fires, and I think we are long past due a complete rethinking of how our “911” services actually operate.
Trump is apparently an insensitive moron for putting the blame on forest management, but I don't understand why the spotlight isn't already on them since things aren't improving year over year. PG&E power lines caused multiple fires last year as well. I wonder if raking and cleaning and doing things underneath power lines would have helped. That may be simplistic logic, but the logic behind recent policy seems even worse and this article makes it sound like they are still figuring out the basics. I am not sure why that is the case.
For example, what if you had the surveillance coverage to detect a fire before it covers, say, even 1 acre, and if you could respond with a tanker of Phos-Chek (or whatever) within some number of minutes?
Aside from actually achieving the detection ability for alerting on a new fire, and having the equipment and personale distributed and on-call for rapid response, I am guessing the biggest challenge is “false positives” in the sense that you detect a real fire which is relatively tiny, but have no way to determine if it has the potential to develop into a catastrophic fire. In the long tail of forest fires there are perhaps 100,000 detectable fires, so that’s a lot of response.
This goes against precisely the thesis of TFA and perhaps the counter-argument is that strictly controlling fires will just allow fuel to pile up to the point where it is no longer controllable — like a dam holding back the floodwaters until it is finally topped.
The biggest problem I have with the “let it burn” argument is just the macro level stats. We have about the same number of fires, burning about 10x the acreage as they did in the past. We are already letting these fires, intentionally or not, burn a lot more acreage and the number of fires is roughly constant.