Some good reading on camera noise (from guys who know what they are talking about!)
Camera Noise and Temperature Tutorial
For images under discussion here in PF, note that shot/electron noise is vanishingly small on a per pixel basis - as suggested by their first picture. For exposures less than, say 0.1 second, the expected shot noise per pixel is a tenth of an electron or less - only a few pixels out of every hundred will seem to have a “signal” - as opposed to the readout noise.
For images taken with a lens cap on, there will be no photons, so photon shot noise will be non-existent.
That leaves readout noise (noise generated in the electronics that converts charge from a pixel to some signal level) as the dominant source of noise here.
That should be pretty much independent of exposure time, I would think!!
Noise, Dynamic Range and Bit Depth in Digital SLRs (By Emil Martinec)
His quote: “In the real world, the raw level does not precisely reflect the photon count. Each electronic circuit component in the signal processing chain -- from sensel readout, to ISO gain, to digitization -- suffers voltage fluctuations that contribute to a deviation of the raw value from the ideal value proportional to the photon count. The fluctuations in the raw value due to the signal processing electronics constitute the read noise of the sensor” sums up where the read noise comes from.
When I find some time, I will see what Rawdigger can give me for some pixel-by-pixel data, so I can calculate something similar to Martinec’s figures 1 and 2.