is there any advantage in logging, say 1 or 2 second epochs instead of 5 when a station is marginal (ie obstructions and trees nearby)?

This question assumes the base and rover log at the same interval.

I would have to use 1 second in order to sync with CORS (that logs at 5). Meaning if I log at 2, then I only have a common epoch every 10.

It seems to me that increasing the data rate is only going to increase the quantity of bad data without changing the bad/good ratio. On the other hand, if you have the storage capacity and don't mind the increased download times, it won't hurt anything, either.

There is no advantage to logging at the 1 second rates over the 5 second rates in harsh environments.

In fact, the downside, as Jim pointed out is that you are getting more bad data and much faster.

You are also filling up your memory much faster and creating larger files to manipulate.

That is my experience.

Deral Paulk, PLS OK

The cure for a poor environment is more time. Give the satellites time to move significantly and peek through the cover. Then a little later on it can again peek through. Solutions can squeeze out decent values if you have several short intervals of good data among all the garbage. I have seen solutions I could easily live with from files where I could not see a solid line in between all the ?s, !s & Xs in the raw data.

Paul in PA

If the bad data is caused by multipath alone or low horizon sv's, local interference like a transformer, I agree that more is not better...

In heavily wooded environments like here in Maine, White Pine has thin wispy needles that don't affect signals much BUT the branches themselves certainly block signal. Spruce and fir have flat branch patterns that act as horizontal stratification and layer themselves, with small holes between the sprays. Hardwoods have horizontal leaves that again act as flat plane signal blockers, especially big flat maple leaves.

I have taken 1 second data and then parsed it down to 2, 5, 10, 15, etc. and processed the files to see where the fix versus float lines are and where the network solutions fail, too. Compared with ground truthing with total station, the 2 second data (in AS,DOS Prism, AOS (geogenius), and Reliance software) all seems to work best for my canopy conditions.

In the open or in multipath environs I probably agree...In closed canopy 2 sec reigns for me.

In a mostly related vein-

I have had units set up on high, open peaks (no trees, no power lines, nothing, i.e., clear views to the horizon), and still see a loss of lock here and there in the raw data views. 1) Why does this happen, and 2) if a person were using a shorter recording interval, would the unit regain lock quicker? (I'm thinking of a kinematic situation here.)

Skylar

Lock is regained at the same rate, whatever the data recording interval, because despite the recording interval, the receiver is constantly analyzing the GPS signals.

Paul in PA