Newbie question re: DOP/satellite position/surveying

Given that you see 8 satellites from each position, there is a high probability that you will have at least 6 common satellites between the two. The key to post processing is to have a common observation time between the two points within which time 99% of the time the same common satellites are in the sky all that time. In 20 years of GPs observation I have yet to worry about an obstruction diagram. A very few times I had such poor commonality of satellites that I achieved more precise results independently solving each GPS point from available CORS and discarding the filed vector. You have to learn enough about your available satellites in view to know when to extend your observation times to ensure satisfactory results the first time around. I have always relied on my instrument reported DOP and skyview to give me enough information for that decision. High DOP based on satellite quantity is useless if you do not have good relative positioning in the first place.

Paul in PA

Unless points A and B are several hundred kilometers apart or more, they will have common satellites for the most part. The satellites will rise and set at different times, but will be overhead at the same time at each point. And on a baseline that long your observation time would be several hours - at least six, probably more - so you'd have plenty of data. If you think about it, the satellites are orbiting at around 20,000KM; the area that can see a satellite at any given time is probably a quarter of the earth's surface.

Regarding obstructed points - my philosophy on that is that if you're occupying a control point with static GNSS then you can probably set it at a location with clear sky. I would only include a point with any significantly obstructed sky in a static survey if it was absolutely necessary to do so, and then that point would get lots of attention - long, independent, redundant observations. Anyhow, these days the processing software (the one I use, anyhow) is so good that it handles data well that would have been a problem, say, 15 or 20 years ago.

Even DOP really hasn't been a problem since the constellation reached it's current capacity. If you're using GPS only for your processing it doesn't hurt to take a look at it; bear in mind that it doesn't change from day to day beyond occurring 3:56 earlier.

**Edit** OK so before someone finds it necessary to correct me a quarter is probably a bit of an exaggeration.?ÿMore likely?ÿabout half that but you get the point.

This graph is using an almanac that's a couple months old but close enough... as you can see, the GPS-only PDOP never spikes above 4 and only spikes above three twice, very briefly. Of course, this is for unobstructed sky.

Thank you all for the very helpful information.

Mission planning and obstruction diagrams were a regular thing back before the GPS constellation was fully operational (mid 90's).?ÿ ?ÿThese days, with GPS long since filled out, GLONASS fully operational, and other constellations coming on, you would only think of doing such things for the most marginal of sites. If you have any kind of sky at your sites, the probability of not getting a solution at all is pretty low.?ÿ The possibility of getting a poor quality solution at a marginal site is much higher of course.?ÿ ?ÿ

Good practice:

Park your base where it has 100% sky view. (no obstruction higher that 10 degrees up from the horizon.)

Best is a bald hilltop. (good radio on a hilltop)

If you had a fisheye camera, that spanned horizon, to horizon, and took a pic at base, and at every sideshot, and you COMBINED the images, you would have a pretty good idea of how gps "sees" obstructions.

Pines obstruct more than oaks.

What brand, and model gps are you using?

N

?ÿ

?ÿ

Posted by: Norman Oklahoma

Mission planning and obstruction diagrams were a regular thing back before the GPS constellation was fully operational (mid 90's).?ÿ ?ÿThese days, with GPS long since filled out, GLONASS fully operational, and other constellations coming on, you would only think of doing such things for the most marginal of sites. If you have any kind of sky at your sites, the probability of not getting a solution at all is pretty low.?ÿ The possibility of getting a poor quality solution at a marginal site is much higher of course.?ÿ ?ÿ

I agree with Norman, GPS has been pushing passed all the old paradigms.

We used to carry sheets on a clipboard for each static control set-up, it included a obstruction diagram and a place to write down HI's in ' and m, start/stop times and a lot of other info, basically a hard copy spread sheet.

Over time that info got recorded in a much shorter version on a page in a field book. I can't remember one time where the obstruction diagram came in handy.

But today, if you have modern GNSS it's not very useful at all, if you have a 4000 series Trimble GPS you may want to do things more old-school. Then as now good GPS control should be as open to the sky as possible.

Thank you all. I should have introduced myself ƒ?? I am student (an older student) ƒ??.

The instructor set up three base stations (?) in clear areas (for a few hours) and we were instructed to make four measurements (15-minute intervals each) in non-clear areas. We needed to draw obstruction diagrams on paper for the four non-clear areas (oak tree canopies, redwoods, buildings on sides (a campus environment), chain link fences blocking the view) and input them into the Topcon occupation planning software. We used Topcon HiPer Ga GNSS receivers (using both GPS and Glonass). (We did have to create the "hard copy" spreadsheet).

I was just trying to conceptually understand the post-processing math, as it seemed we were only drawing obstruction diagrams for the four rover (?) stations (we did not have RTK). But really, what it seems you need to know for the post-processing to work is a common set of satellites between the base station and the rover.

I think I can see from your responses that the probably of a satellite misalignment (A sees completely different satellites than B) is essentially nil. It can happen, but itƒ??s very unlikely, for the reasons you all mentioned (# of satellites, distance to earth/size of earth, receiver quality) . So as long as one pays attention to each individual measurement to have good visibility or wait a longer time, the post-processing software should be able to improve the measurement accuracy, as designed. At least thatƒ??s what I am understanding.

I've been working with GNSS since the early 90s.?ÿ Back in the day there was a lot of effort put into pre-planning a static GPS mission including the collection of obstruction diagrams for each station.?ÿ I would cruise the project area recovering or setting points as required and then preparing an obstruction diagram and to reach description for each.?ÿ The normal method was to use a clinometer but I did try to develop a system using a fish-eye lens to collect the data.?ÿ It was actually kind of fun.?ÿ The basis technique was to create a combined obstruction diagram for each occupation.?ÿ Sometimes that meant 2 receivers while others it meant 4.?ÿ As pointed out, unless you are covering very large areas (read hundreds of kilometers between stations) then you can figure that all of you station will have the same SVs in site if it were not for the obstructions.

Those days are long since past, with just a full constellation of GPS I rarely have fewer that common 6 SVs in site.?ÿ I do not think I have done an obstruction diagram in years now unless it was required in a contract.

Mighty, I will say that I do still record the HARP and start time for each static/control observation.?ÿ Hell, if the project is purely GNSS I still use an observation log form.

?ÿ

We record monument information (instead of a diagram we take a photo), start/stop times, PT#,?ÿHI in feet and meters, receiver number, point on the receiver the HI is measured to, tribrach info, we start the receiver with a data collector and imput the HI with the meter # and an m behind it which will convert it to feet as you watch?ÿfor a check. Probably the things that have been most useful are the receiver #, PT#?ÿand the feet meter measurements. Post processing errors are usually imput problems more than GPS data. It's important to track PT#?ÿwith the?ÿmonument descriptions.

We record all the usual stuff on an old school log sheet, plus on points we set we record at least three ties to nearby objects. Start / stop times, height, filename, point ID, a checkbox stating that the setup was verified to still be level and plumb at the end of the session.

Planning for obstructions was definitely important at one time; in the early 90's you also had to plan for satellite availability and DOP, and schedule your sessions accordingly. The problem then was that your control was existing monumentation that had been set before anyone even dreamed about GPS and clear sky was not a consideration. Finding monuments that were suitable for GPS control in an area where you hadn't surveyed before could at times be the hardest part of the job. One of the nicest things anyone ever did for me was when the LA NGS advisor presented me with a bound copy of the Louisiana HARN, that was worth its weight in gold at the time. Thankfully abundant CORS stations and the internet have changed all of that, GPS surveying is a whole lot easier now.

Posted by: Lee D

redundant observations. Anyhow, these days the processing software (the one I use, anyhow) is so good that it handles data well that would have been a problem, say, 15 or 20 years ago.

?ÿ

Which software are you using Lee?