I'd think that with all the technology we have these days, early detection of fires wouldn't be such a big problem.
So, since my degree covers the technology most likely to look for fires over a region, I'll tell you what I know.
Remote detection of anything is based on the energy it puts out and the medium of transmission. We can measure earthquakes from far away due to the energy traveling a long distance through the ground, and likewise can surmise that a tsunami has been generated before anybody sees it because the seismic waves travel faster than the ocean waves (and therefore we are alerted to them before they hit, and can possibly find them and estimate their impact). Fire puts out a lot of energy, but most of that energy travels straight up. It gives off visual light which can be detected laterally, but due to obstructions (forests, buildings, and the curvature of the Earth), detection methods from the ground remain pretty low-tech and are essentially manned watchtowers with radios (which I do believe California has; the darling indie game Firewatch cast the PC as a ranger assigned to watch for fires, although in another state).
If you can't automate detection from the ground (or underground, for what that's worth), then your remaining mediums are air and space (which, depending on how you classify different layers of the atmosphere, can overlap).
--------
Air: Every time you put up a drone or manned aircraft to collect data which can be analyzed to find fires, you're paying costs for operation. If you don't find a fire, you've thrown money away. I don't know how California chooses to balance that. Perhaps they identify high risk zones and fly them once a week. I doubt they fly them every single day, the cost to tax payers for that much area would be monstrous and all government institutions never have as much money as they want to fulfill their mandate.
Space: there are several satellites in orbit equipped with remote sensors to gather Earth data. Due to the mechanics of cameras, cost-effective space-borne sensors typically take images which are much lower resolution than from cameras made for drones or manned aircraft. Therefore, they may not pick up the thermal infrared energy from the beginnings of fires, as whatever energy the fire is radiating is averaged out across the area of the ground that a single pixel of a satellite image covers. The other problems with satellites is that an individual satellite may not return to the same area of California for a month or more, depending on its orbit; the orbits are designed to be somewhat less than 24 hours so that they can scan an adjacent line of the Earth every day until they come back around to the same spot. There is such a thing as satellites with geosynchronous orbits, which means that their orbit matches the Earth's rotation and therefore remain in the same spot relative to the surface (for example, always over California). However, these are mainly communications satellites; remote sensing satellites pay for themselves by selling their imagery, and the costs of satellites are such that a global market is preferable to a regional market, even if the global market only updates data once a month (very few companies and agencies require even monthly updates to regional imagery, let alone daily). I don't believe there are enough remote sensing satellites with the sensors to detect fires in operation to provide daily coverage of any particular area of Earth even if California decided if it wanted to pay for all of that.
--------
As a quick aside, something I've glossed over to this point: just about any visual light sensor will be able to find a fire by seeing its smoke. However, to see the fire itself, you have to see through smoke and foliage. Therefore, seeing the actual center and extent of fire to target efforts towards typically relies on thermal infrared, which, to put simply, measures the radiation of heat, and is not a capability of all sensor platforms. IIRC thermal IR sensors are typically lower resolution in general compared to visual sensors (in this context sensors are synonymous with camera), due to the mechanics of those kinds of sensors, which adds difficulty. This is multiplicative with the principle I mentioned in the previous paragraph that satellite sensors across the board are lower resolution than air-based sensors.
As another aside, there
are fine resolution satellite-based sensors. These are either:
A) military satellites
B) espionage (spy) satellites
C) extremely expensive to make and charge very high fees for their data
There's a lot of overlap between A & B, and neither are typically available to civilians (their data
can be given to local governments in a state of emergency, as part of coordination between military and civilian emergency management, but that's only after an emergency is ongoing).
And as a final aside, what I've said about satellites applies to the US. Satellites are the provision of powers with space programs. The US has the largest and most robust overall satellite system in the world, followed by Russia, with China, and all European sources (including British Commonwealth countries) combined, occupying a third level, and Japan and India trailing fourth. The rest of the world collectively have very few satellites (and most of what they have are devoted to military or communications), and must make business and political arrangements with countries that have greater capacity and options.
--------
So, indeed, there are cost and logistical issues with fire detection which is simultaneous with fire generation. There could be other branches of technology outside of what I'm educated in that are used to detect fires, but so far as I know, fire detection remains mainly a matter of manual reports from either rangers on fire watch or civilians, and they are more likely to spot larger smoke plumes than smaller ones. Air and space methods are most typically used to track known fires and detecting new ones in the same vicinity, rather than proactively searching them out.
Can we have fire detection which is simultaneous with fire generation? I think that we have the technology capable to do so, or are close to it. There's a good bit of research right now into high altitude, geosynchronous balloons, able to be launched with small, cheap rockets (such things are already in the province of hobbyists) and hang in the upper atmosphere without maintenance for long periods. I think that a design for such a balloon with a fire-detection sensor package is within reach of the next 10-20 years of technological development, if not already, and such a design could be mass produced to create a network spanning the length of California. The implementation of such is mainly up to a team putting in the research and design hours to invent it, and the willingness of the state of California to purchase the requisite number of balloons and hire on a staff to keep an eye on them and analyze their data.