In your "Just the Tips / Foggy Morning in Vancouver" shot, I do see three dust spots (upper left area) on the sensor, but I don't see a "noise" problem. One would have to be crazy obsessed to think that image has a "noise" problem. I do see three hot points that look like they may have been light bulbs in the town below. But they don't add anything so I'd probably retouch them out.
In your "above all" shot I do see some graininess in the clouds in the foreground (and if I really squint I can see a very very tiny amount in the background - but that's a stretch to see it.) This may be caused by your processing technique.
In any case, you can knock back noise in software. I've experimented with different tools but never thrilled with the noise handling performance of Photoshop or Aperture, so I started trying various noise handling plug-ins. The favorite that I finally settled on is Noiseware Pro (imagenomic.com). Noise Ninja is also popular but I felt I had a tiny bit more control with Noiseware Pro. BTW, you can get this as a stand-alone app or as a "plug-in" and the plug-ins are available for Photoshop, Lightroom, or Aperture. Some cameras generate more "chroma" noise, some more "luma" noise. Often noise shows up more in shadowy areas and less in highlights. Before I go too far, I should mention the basics of what de-noising is.
In high ISO and especially high ISO / long exposure images you get more "noise" in the image. This is randomness that causes some "pixels" to read more light than others ... or even false color that didn't really exist. But it's usually at an individual pixel level. (Also heat generates noise. A sensor which is physically cold will generate much less noise than a sensor which is physically warm.)
To de-noise, the software analyzes each pixel and compares it to the values of surrounding pixels... but it does look for edge-transitions, etc. If it sees what seems to be a random isolated pixel with a color value or light value that noticeably departs from adjacent pixels, then it assumes it's "noise" and it "averages" it's value to the value of the surrounding pixels. This makes the noise stick out less... but this "averaging" can also have an effect on genuine detail in the image -- mistaking fine detail for "noise".
Basic de-noising simply averages out everything -- you don't have much control. It will knock back the noise, but it also noticeably softens the whole image.
Advanced de-noising software is a bit more clever. It can help with noise analysis, but it also allows YOU to use your eye to "tune" it's performance. Most cameras have stronger noise in shadows and less in highlights. The software allows you to "tune" this so that it doesn't do much (if any) de-noising at highlights (so you don't have a loss of detail where it's not needed) and can be more aggressive with noise in shadows. It can also be more or less aggressive about certain types of noise.
Most of these de-noising plug-ins have a free trial. You should probably take advantage of that and try them out before you decide what you like best.