Wind is less liable to generating forest fire than high air temperature, low humidity and a long drought. But the wind is most responsible for the spread of forest fires. Generally, the higher speed/velocity of the wind increases the spread of forest fires. Also, unfavorable wind (speed and direction) complicates fighting forest fires and sometimes prevents shutdown. One analysis of forest fires, depending on atmospheric parameters, can be found in the reference below.
But in my study area bamboo mixed forest are common during summer high temperature and low humidity with high wind velocity dry bamboo pole collusion ignite fires in forest.
I couldn't agree with you about first two sentences. Could you please check this paper: http://thermalscience.vinca.rs/pdfs/papers-2014/TSCI141103150R.pdf
Only in the USA, according to official data, the number of forest fires is between 52 000-92 000 per year. For me it is impossible to explain these phenomena by human activity (intentionally or unintentionally). Similar absolute values one could find for some other regions: Regarding the region of the Mediterranean it is particularly necessary to point that: “A recent regional situation analysis published in the frame of the FAO Global Forest Fire Assessment 1990-2000 …reveal that the average annual number of forest fires throughout the Mediterranean basin is close to 50 000... (Goldammer G. J. 2002: Towards International Cooperation in Managing Forest Fire Disasters in the Mediterranean Region. International Forest Fire News/GFMC No. 27, p. 81-89).
In short, if we talk about lightning like possible cause for ignition of forest fires I could recommend next source: From 1990 to 1998, over 17 000 naturally ignited wildfires were observed in Arizona and New Mexico on US federal land during the fire season of April through October. Lightning strikes associated with these fires accounted for less than 0.35% of all recorded cloud-to-ground lightning strikes that occurred during the fire season during that time (Hall L. B. 2007: Precipitation associated with lightning-ignited wildfires in Arizona and New Mexico. International Journal of Wildland Fire 16(2), 242–254, DOI: 10.1071/WF06075).
Satellite observations, according to my opinion, show geographical positions of "hot spots", but they couldn't be the proof for human activity or lightning like potential explanation.
There is really a lot of sources for this kind of discussion, but I am afraid that it needs a lot of space. Anyway if you are willing we can continue our communication via e-mail.
as a weak mechanism I can think the wind as a bias in the fire propagation, like in the biased in the reaction diffusion process.
as a strong one, that is, increasing the reaction rates (fire+matter) itselft, the propagation front should have velocity higher than the case without fire. if you can compare those situations experimentally you should be able to identify the role of fire for your particular case. take a look at theoretical models for this type of problems, it might help also.