Whenever I seach for radar implementations, independently of application (weather, military, autonomous vehicles, speed limiters, etc) it always comes in many flavours: huge antennas, small antennas, high power, low power...but never low frequencies. They are offen in the GHz area. Why one’s not able to create a radar using some MHz? Is it related to directivity? If so, why MHz cannot be as directional as GHz?
To a first order approximation, you get directivity by making the aerials large with respect to the wavelength.
It is much easier to make a moderately directive antenna at say 24GHz then it is at 24MHz where the thing would need each dimension to be 1000 times as large for the same beam pattern (Or course you may not use the same design, but the point stands). In addition modern electronically steered beans really need the elements to be ~1/4 wavelength apart, so again MUCH smaller in the microwave bands then at VHF.
In addition while narrow beams help with cross range resolution, resolving distances depends strongly on system bandwidth and 10Mhz of bandwidth is MUCH less of a big deal at 24GHz then it is at 20MHz (where it would be a full octave).
Further, to have a decent cross section targets need to be larger then half a wavelength or so, ideally much larger, 24MHz is a wavelength of ~12m, so you probably want quite a bit higher then that.
Now one place low frequency radars were very much a thing back in the day were the massive over the horizon arrays used to detect ICBMs coming over the poles, they were enormous machines with multi megawatt power levels and at least in the Russian case used to like to frequent the HAM bands, much to the disgust of shortwave radio geeks everywhere. Search for the "Russian woodpecker" for some good video of the installations.
Any measurement or imaging system using electromagnetic waves has trouble resolving features smaller than the wavelength of that feature - smaller features "fall in between the waves" and/or create erratic effects like aliasing. This is also the case with microscopes, cameras etc. - a normal optical microscope will not resolve features much below 1 micrometer no matter how many magnifying optics you stack on top of each other.