You're right, I'm not measuring nearly enough.  For example, my depth only goes to about 1m, I don't realistically have a way of digging deeper than that.  I do have wind data, rain data, but am missing sky data right now. The graphs in grafana are a bit too unreadable if I throw all the other variables I measure into them, so alot of this will be gathering long periods of data, and then going back to the graphs and looking at them vs different factors, like humidity, etc etc.

One thing about doing this in Phoenix, and with the rather dense structure of the ground, is that there is virtually no water penetration at depth, so between 3" and 36" it's pretty much bone dry 365 days a year.  IMHO this makes my set of variables a bit simpler, as the material's properties aren't changing day to day.

I think my short answer to the above is, I'm aware of most of what you said, but alot of this experiment is more along the lines of visualizing it myself, in an environment I understand.  I've read papers and seen graphs of data, but it's always in a climate that is far different than mine, with soil and rain conditions that are far outside my daily norms.  The idea of even having "soil" is insane to me.  :)

I don't expect to make a scientific breakthrough here.  More that I want to slightly understand the mechanism better in my own back yard.  I think what I'm really curious about here is kind of similar to the concrete heat island effect.  I live more or less on the outskirts of the city, so the quantity of concrete in my area is far lower, so I believe my air temperature is more influenced by ground than concrete.  I'm wondering if I can predict "record high" type days, by looking at how the ground temp is moving.  IE, where is the point where the overnight cooling simply cannot happen because the ground is buffering it up.  Can I identify that?

I'd be glad of any references you have though.  Part of this is gather what data I can for awhile, stare at it, and figure out what I'm missing, then figure out how to grab that data.  The problem with weather projects is that your first data cycle takes a whole year.  :)

I'm not sure what you are doing. I genuinely mean no offence by my ignorance but I would like to try and understand better. This is what I understand of soil temperature. The ground temperature at depth (say 20m) tends to the annual mean dry air temperature. The surface temperature tracks the daily fluctuations, including effects of insolation and evaporative cooling. As you go deeper the seasonal temperature oscillation begins to dominate the diurnal cycling, with a time lag that is proportional (iirc) to the depth. So, for example, at 3 metres depth in summer the soil might be at the coolest temperature it ever reaches each year. The amplitude of temperature swing diminishes and eventually you are at the mean annual air temperature. 

I also understand that heat will flow from hotter to cooler substance, where it has a mechanism to do so. So if the air temperature drops below the surface soil temperature, then yes, it will be heated by the soil. This is a basic fact so I'm sure you are trying to test something more nuanced. But there are several other heat transfer mechanisms to consider which will impact the mechanism you are focusing on. To illustrate, if you have clear skies at night, the soil will radiate significant heat into space (the radiant temperature of sky is very low) and drop in temperature. This can easily drive the soil temperature down and undoubtedly affects the direction and degree of heat transfer between soil and air. Another example is whether the air is still or moving (wind) and what speed it is at the surface, where it came from (and the temperature /humidity there) and what expedients to turbulence or natural convection (like buildings, ridge lines) there are. 

There are many factors to consider. I don't know what you know about the topic already but it seems from my reading of the tests as currently written up that you are not measuring enough variables to be able evaluate the contribution of heat transfer from soil to air in the scenarios you are interested in. Would be glad to hear your thoughts, especially in the case that I have misunderstood this project! Will try and be helpful if I can, from my admittedly old information on the topic. I probably have some references to useful papers for this kind of investigation. Good luck!