More success on the ULF 101km (2.97kHz) band

Today I heard that Stefan DK7FC's 100nW ERP signal on 2.97kHz VLF was received in a 212uHz bandwidth at 16.8km. He will soon be trying to receive this radiated signal at 30km. I wish him every success.

Hi ULF friends,
This is the summary of my portable RX experiment in 16.8 km distance to my transmit site. It is into the far field of the 101 km RF wave.
A 100 nW ERP signal was sent and recorded for nearly 2 days.
The recording was started on Saturday, 30th of July (14:58 UTC) and was finished on Monday, 01st of August (12:25 UTC). The transmission was a carrier sent on 2970.000 Hz, it was shifted by +2.5 mHz at 8 UTC on Sunday and again shifted by -5 mHz on Monday, 8 UTC.
In the post processing of nearly 30 GB wav files the signal was successfully detected.
Due to the exceptionally high QRN levels (even for the summer time), the SNR was lower than expected, however the traces and frequency shifts are clearly visible.
Spectrogram of the signal in 212 uHz FFT bin width: https://dl.dropboxusercontent.com/u/19882028/ULF/17km212.jpg
Spectrogram of the signal in 47 uHz FFT bin width: https://dl.dropboxusercontent.com/u/19882028/ULF/17km47.jpg
A wideband spectrogram giving an idea about the natural and man made noise conditions as well as the frequency response of the tuned loop antenna: https://dl.dropboxusercontent.com/u/19882028/ULF/17kmwide.jpg
Wideband spectrogram zoomed to the spectrum of interest: https://dl.dropboxusercontent.com/u/19882028/ULF/17kmwide3.jpg
The best SNR shown is 12 dB in 212 uHz and 17 dB in 47 uHz.
During the post-processing procedure it happened that SpecLab has plotted a few pixel twice. I don't know the reason.
It looks like i am now seeing the natural background noise on 2970 Hz even in quiet periods (the faint vertical traces). However the man made noise is still playing a role in this QTH.
Further informations can be found in the older emails below.
73, Stefan
PS: A next experiment is planned in a few weeks. The goal is 30 km distance. With lower QRN levels, a better tuned antenna and higher power (already running!) this should give even better results.