I added a doppler search mode. I had originally calculated the doppler effect for a very low frequency and did not realize the doppler effect is more profound the higher the frequency. This changed things. Since originally, I was thinking I didn’t have the frequency resolution to do a doppler search.
I also worked on trying to subtract the transmitted signal from the received signal. I am suspicious that this will not improve the correlation, but I could be surprised. It is showing interesting results.
The doppler results are confusing at best. I get results that defy my expectations and may be bogus. When I subtract the original signal, I get what I feel is slightly improved resulted.
The above is some output from the doppler search function. A dark blue indicates the chirp was shifted lower in frequency by moving away from our observer. The greenish is in the middle indicating no shift and the red indicates the chirp had the strongest correlation magnitude for a higher frequency shift.
It is obvious that at around the 4.32km point the function enters into noise, because there is a nice balance and mix of various colors. There may be an underlying light blueish tint! I’ve seen it in other zoomed out views indicating an overall slight movement away from the observer for the particles of water vapor and such forth that reflect the signal.
The major problem is that the output is not changing enough to be discernable even with changing weather conditions. At least at this point I’ve seen no changes that correlate with the weather. I also find it difficult to make sense of the output in terms of perhaps more stationary weather or atmospheric systems.
The reflectivity, above, shows a lot of reflection at the first sample then it skips a few and has two major samples of energy. I also see some other places at 6.10km and 39.18km where it seems it wants to create a band. It is also evident the signal abruptly discontinues around 2km because it is mostly noise except for those two bands which might just be artifacts of processing.
The phase delta is interesting. It shows what is expected from the reflectivity view at the beginning with a few samples of various phases. Then there are some places that do not appear to be noise and are not continuous.
I have been tired of finding output that stays continuous. To have something that shows some variation over time is very pleasing. I expect that since the weather varies over time then so too should the output if there is a correlation. It is also possible the variation is hidden under the stationary process output.
I’ve subtracted the phase overall average and average per column. The variation is more visible now. I cannot find much to be excited about. It is either very stationary a in confined to a single column or too sparse and noise like.
Conclusion
I think what is happening is that the output is primarily showing the transmitted chirp and any reflections are just too weak to be observable using my current hardware setup.