Over the next three months I need to determine sub-meter lat/longs for multiple points at 100 sites spread over Northern California. As non-survey staff will be doing some of the observing, I need simple to run receivers. Beyond this initial 3 month survey we’ll have the occasional additional site needing lat/longs but otherwise haven’t yet identified further uses for GIS level GPS equip. We do have CORS stations within 50 to 150km and Coast Guard Beacon coverage for most of the work area.

Along with this work, I need to venture into survey grade GPS. Given my needs(and my “just getting started” budget) can anyone recommend GPS equipment? I keep wondering if I can purchase PM2’s or Locus units and do 1 receiver surveys using CORS. That way I could cover two "sub-meter sites" at any one time or I could use the receivers together for survey grade measurements. What kind of results would 1 L1 receiver and CORS yield? Is distance to the CORS site a big issue?

If I’ve used some correct terms here it’s from reading your posts. Thanks for the help.

Chris Johnston LS

You cover a lot of ground with your questions Chris. I am fairly green at GPS too, maybe just a bit browner shade of green. As a matter of fact the more I learn the "greener" I feel.

But to some of your questions...

Distance is definitely a consideration with respect to GPS surveying. As a matter of fact it is likely the greatest consideration (along with satellite availability) envolved in equipment and procedure. This is due (as I understand it) to the variation of atmospheric conditions that delay the satellite signal. Points to observe near each other do not have as much variation as points further apart. The Locus and Promark2 systems you mention are L1 only and do not support "sub-meter" work. L1 is excellent for survey grade (1cm or better horizontal) with points no further than 12 miles apart. (12 miles is somewhat arbitrary, it is possible to go a little further, but results are not as reliable.) I would suggest looking into perhaps leasing a system specifically designed for submeter GPS work for the job at hand, and then looking into your future surveying needs seperately. Even though a nice "fit-em-all" would be less complicated, you may be sacrificing the benefits of having taylor made systems.

Hope this helps,
Shawn

By far the most relevant issue to your inquiry is site suitability, i.e. openess, vegetation, valleys, hills, under power lines, etc.

To give my opinion on one of your questions. CORS would work but distance is definitely a factor for L1. Based primarily on intuition I wouldn't expect to get more than 30kms reliably, but I don't know what would happen if you just want 'sub meter' with survey grade equipment. You might be able to push out farther and get a sub meter solution, but how reliable is it really?

If you are going to use untrained personnel hopefully points are accessible as you can expect to have to re-observe some percentage.

Chris Johnston wrote:

Over the next three months I need to determine sub-meter lat/longs for multiple points at 100 sites spread over Northern California.

Sound like you might be planning to do some GeoTracker work. The PM2 should work fine for this when conditions are decent (pay attention to solar weather) and appropriate procedures are employed. I've done some testing with the it and find that I can get submeter (generally in the 3 decimeter range) using 1-hour sessions with 3 well-distributed CORS out to 100km or more. Since the PM2 is L1 only, you won't be getting fixed-integer solutions, but that's okay for GeoTracker purposes.

A month or so ago I requested that GeoTracker add the PM2 to its list of approved equipment, and this has been done.

If you're planning to locate more than one feature at each site, or features that are in satellite-hostile locations (like monitoring wells near buildings or under canopies), you'll probably want to run at least 2 receivers so that you can establish control for locating the features with conventional equipment.

The PM2 solution isn't as elegant as having dual-frequency gear, but it's a damn sight more affordable.

For those unfamiliar with the term, GeoTracker is a California statewide GIS for environmental data. Recent legislation allowed the regional water boards to require that all groundwater monitoring data be georeferenced at the submeter level, so there are a whole lot of gas stations across the state that are now getting a visit from people toting GPS receivers.


Modified By Jim Frame on 11/6/2001 at 8:14 PM

Explain your procedure a bit more in detail. Are you downloading CORS data from 3 stations in RINEX format to process with each 1 hour session for the PM2? It's a really interesting concept given adequate CORS densification in the work area. In my case the nearest CORS is 65 miles, the next 120 miles or so. Not much good here, but maybe some day we'll see more stations.

Curious

J.D.

I'm going to address the sub-meter project...

I'll try and keep this generic...But the two (sub-meter and centimeter) really don't mix well..Most instruments are made for a purpose, and generally do one thing better than the other...
I'm going to mention the T-word, but don't take this as an endorsement, as I'm sure Ashtech provides similar results...We use a 12 channel Trimble PRO-XRS with the John Chance Omnistar satellite corrections. It has proven reliable across the country with sub-meter second by second results...Find an Ashtech Reciever that can use the Chance Omnistar and you can accomplish all the sub-meter positions you need.
Better yet..It's real time, so you will not have to process anything..We usually take several minutes, since you've already driven to the site, then this is readily justified. Sometimes I have taken several files per location, just to provide some statistical verification...

The Coast Guard Beacons will not provide sub-meter results in all situations...If I was sending results, then I'd be hesitant...Good for navigation, but less than good for reliable sub-meter results...

Ask Phil..We checked it in Lawton on a demo and it was much farther out than sub-meter...

Shawn suggest leasing a system for your project...This is a good idea...Try before you buy...Also, just to insure that your confident in the equipment..Make it prove it's accuracy by occupying a HARN or two before you start the project....

You mention that some of the collectors will not be surveyors (surprised that no one flamed you on that, but I wouldn't have in any case)..
You will need a specific set of spec's for them to follow for these surveys....
Minimum 5 SV's, PDOP<4, ect....

TM

You need to consider Ashtech Reliance for your sub-meter GIS data collection. You could use your CORS data and obtain sub-meter results with some good field practices. Shawn is right on the money, lease or "rent" for your 3 month period is a good idea.

I am the Ashtech dealer in Colorado, contact me privately (if you wish) and I'll help you get hooked up with the right product, and the dealer nearest you.

Chris
Modified By cj ssi on 11/7/2001 at 12:24 PM

Jim Frame has it right, this work is driven by State Water Board requirements. And like he has experimented using L1 survey equipment, I was hoping I could put the equipment expenses for this work towards something more useful to me, i.e. survey grade GPS.

I am a bit confused as to how Pro XR and Reliance equipment can claim to get post processed sub-meter positions but L1 survey equipment apparently struggles to do so. Must be something missing in the equipment claims? How would L1/L2 equipment compare?

Trimble Man: I expected to get flamed. What I didn’t say was that we already know the relationship of the points at most or all sites. We just need to know where they fall relative to NAD83. What we do know will give us a check of sorts on the 3 or more GPS positions we determine at each site. Once we know what procedure is required for each site we’ll decide if non-survey staff can handle the GPS obs. It would be nice if non-survey staff can help as they’re already going to be on-site for 4 to 16 hours.

Half of your responses so far direct me towards Pro XR and Reliance type equipment. Should I just give up on somehow fitting survey grade equipment into this project?

Thanks. Chris Johnston

Survey grade versus 'resource grade' operate in two different regimes. People who have survey grade stuff are usually surveying and the manufacturers don't necessarily design the software to operate in the non survey regime.

I didn't discuss the other options because you indicated you wanted feedback on survey grade solution.

Jim Frame gave you his experience which seems to indicate it does work. It would be easy for some people to check by taking existing sessions agains far away CORS stations versus the HPGN or nearby CORS that they were using originally.

Ashtech's AOS software would let you tailor a solution to use float solutions and a host of others for either or both L1 and L2 but I don't recall if AS software will let you do that.

Survey grade software is focused on getting that solution down and so it may take some experimenting and thought to figure out the optimal way to use it to get results outside those narrow expections. It could very well be quit easy, and Jim's experience is one indicator that it might work fine.

I am not familiar with the omnistar system, but it does seem odd that it would be that much better than other differential systems as it seems to be using a similar technique as WAAS, however it looks like they have 3-5 sites right along the CA-OR area so they can probably be giving pretty good corrections over that area.

Anyway that is tangential to your quest as to whether survey grade receivers (and software that comes with it) can accomodate processing in the sub meter versus centimeter level environment.



PS: AOS Ashtech Office Suite (geogenius) I assume ashtech is no longer selling.

AS Ashtech Solutions, the software that is easier to use but less flexible and confusing than AOS and is current with Ashtech Survey Grade L1 and L1/L2 systems I think.
Modified By jerry wahl on 11/7/2001 at 4:15 PM

Another consideration is that GeoTracker requires positions to be referenced to a minimum of 2 HPGN (HARN) NAD83 points. That might be tough to do with the PRO-XRS approach.

Since your staff is going to be onsite for a few hours anyway, it seems to me that setting up a couple of PM2s and letting them gather data the whole time they're there is the best approach to meeting the GeoTracker requirements. You'll have defensible data (unlike real-time DGPS, which leaves you with little but coordinates) and survey-grade equipment for use on other projects.

I tried to dilute it(the Trimble reference) and expand on the Omnistar system...I'm sure that some Ashtech system can use the Omnistar corrections, which are very accurate...
Jerry made my point,in that survey grade software/systems are aimed at practicing surveyors (who devote a lot of time and training dollars to getting it working perfectly), while GIS type systems are aimed more at the user who needs a good tool, but doesn't want to expend lot's of training time to make it work...I've trained people on the GIS systems in as little as two hours and they gathered very good results(not survey grade, but sub-meter grade) that fit their specs for a project...
Ominstar works much as OPUS, but in a much more real time situation..They take corrections from many base stations and send you the best corrections (in real time) as you gather data on the best possible correction...Most of the time I've seen results in the US Survey 1.0' range...(using HPGN stations as a check)...
Omnistar was developed by John Chance (a very large and very smart engineering company) down in Lousiana...Check out their website..I've visited their control room and they rival any technological program for corrections...I think that we pay about $800 per year for the service...Cheap when you only have to have one receiver.. They work with any receiver that can receive their correction message.

I'm sure that with an L1 only survey receiver, that can easily reach sub-meter at the 100km-200km range...But you'll have to download the rinex, do the processing, ect...(office time and expense)...

You can't do this with the beacons at this range...

TM

I think that the Z-Xtreme can be configured to handle submeter DGPS work it is also probably the best system for tying in long CORS ties. I would think that you could mix a Z-X with a Locus or ProMark and still do on site static work. Supposedly Omnistar is coming out with a system that will allow for a 2 decimeter solution using a dual frequency receiver within the next year. Don't know if that is in your time frame or not, but since you mention it, I think I know what I want for Christmas now...Dad?

Shawn

We use L1 only "survey grade" receivers all the time to generate sub-meter positions, ie Promark2's, Ashtech Reliance. When we do a survey, we throw two recievers on traverse points in the open and just let them chug for a few hours while we are cutting line, doing recon, etc...When we get home we download the receivers, get cors data from the nearest 2 or 3 cors stations (up to 150 miles away!) and process the baselines....This gives us a submeter level FLOAT or better solution on site for future work. We set a receiver or two up the next time we go out and do it again to see how the solutions agree, and 100 mile baselines still agree within 2 or 3 feet 95 percent of the time......When you put the results in to least squares, you find that the float baselines still have validity, just a greater uncertainty in their vector solutions...Looking at the residuals, we still can say with certainty we are getting sub-meter positions out there with "survey grade" units....What am I missing with this discussion about survey grade either being ALL or NOTHING, ie, centimeter or unusable?....bob

My opinion of L1 survey grade receivers for submeter work is based entirely on personal experience with the Locus system (using Locus processor) and a conversation on this board with Brian Ewing on the subject. On our first attempt at pk (failed attempt) using the Locus we basically had almost 20 static positions on points anywhere from 1/2 mile to 3+ miles away with five minutes worth of data on each. Processing the longer lines resulted of course in float solutions. The statistics (error estimates) provided by Locus processor were unreliable at best. While some positions missed only a foot or so some were much worse, with no way to destinguish the good from the bad and ugly. On top of making pk work (which we finally did succeed) I hoped that I could do short, unitialized kinematic observations on points and get submeter results. This too proved unsuccessful (again using Locus processor). When I asked Brian about this, he explained that (as I recall) the antenna was not made to capture the necessary data to do submeter work, as well as the receiver and the processor. It could very well be that the promark2 is more suitable to submeter work (since the antenna is made for carrier and code phase capture). This is a guess on my part, but the fact that you and Jim both mention that this is possible is reassuring.

Shawn

I think processing methods are being confused a little in this thread. Usually "submeter" gps is used in the context of pseudo range code only measurements or carrier smoothed code solutions, so that the rms values ( a radius, not a diameter!) fall within 1 meter...Submeter Ashtech Reliance (code only) and ProXL units fall in that category. The second processing method attempts to fix ambiguities of the carrier cycles so that you can know how many cycles or fractional pieces of cycles have passed since the signal left the SV until it reaches your antenna. Increasing baseline length between receivers until the combination of ionospheric and tropospheric disturbance, receiver noise, and other unmodeled errors reaches the level where detection of a dropped cycle is not possible leaves you at the point of a "float" solution, at best. Float solutions can still yield good vectors, but you need redundancy to sort out the poor vectors from the better ones. For the purposes of this thread, I am stating that float solutions attained with ONE hour or so occupations are for the most part capacle of yielding sub-meter accuracies at long baseline lengths. I am NOT suggesting that short occupations at long distances will do so. I also agree with TM that the Omnistar or Racal geostationary correction services can yield similar results, but I want to see some redundancy in my occupations so I can sign and seal those results as "submeters'...bob

Mr. Bills really brought the topic to the head.Most sub-meter units are typically code correction receivers, while L1 single frequency uses phase based solutions (they do use the code to get initial lock on the SV's)...

Bob-I have never been asked to certify my positions were sub-meter on a map..I would certainly fight this if possible, since some might infer more than sub-meter with a seal affixed..

Since many of the users here deal with cell tower locations. Then let me add one more thing..Using a sub-meter with Omnistar (or Racal as Bill suggested) provides second by second sub-meter work..You can collect alot of data in several minutes and I always plot my all my positions before outputing the final feature in the office..Seeing a nice round shot pattern gives me ample confidence in my data...

TM

Bob,

Thanks for the explanation. I just wanted to clear the air as to why I didn't really trust "survey grade L1" units to provide submeter results. We once did a static session to a CORS 60 miles away at night to our office post. The results were off about a foot as I recall as compared to our coordinates based on HARN. I certainly think that what you describe would work.

Shawn

I just reread the discussion between Brian Ewing, Greg Shimp, and me below about code solutions using the LOCUS and the inferrence is that receiver firmware and the processing engine itself are the promary reasons that quick submeter results are unreliable using the LOCUS system. I would suspect this is also the case for the PM2. As Bob pointed out though, if you have the time to occupy a point you could get a good float solution that would be submeter.

Shawn

The Ashtech Reliance system has been around a while but, because it is an old reliable tool, there are not many who talk about it. It has been mentioned in this thread. It is often neglected in favor of newer bells and whistles. I wish it wasn't so neglected because it is a wonderful tool.

When I bought my Reliance system (I am a Reliance version 1.0 survivor) I bought the decimeter grade package for it. At the same time I bought Survey Control 1 so I could use the SuperCA right along with my Dimensions for L1 static survey work.

At about the time we got our RTK gear I also purchased the centimeter level option for Reliance. Why upgrade Reliance when you are buying RTK gear? Because I knew from the get-go that the radio link was going to be a pain in the neck. I wanted a mapping product that would allow me to escape the radio.

So, let's just take a look at what Reliance offers:

Your choice of submeter, decimeter, or centimeter level processing in the Reliance software.

Your choice of recording intervals so the folks with ProMarkXCM or ProMark2, or Locus, or you-name-it can use the SuperCA right along with those tools for static GPS work.

It will work with the Racal or the OmniStar DGPS services. I used mine with DCI once upon a time.

With the right options and accessories you can use a SuperCA as a real time DGPS base station.

How long a line can you measure from a CORS? How far do you need to go? Test your procedures and find out if you can get far from that base station.

If you want to see a comparison of Reliance work with high precision static GPS work go look at my demonstration project done for the students at OSU-OKC. You can download the shape files from the OSLS web site at http://www.osls.org. If you look at those files you will see data processed relative to a base station located on campus. You will find the same rover data processed relative to a base station located 95 miles away. You will also find the static files we collected with dual frequency Z-12 Sensors.

The proof is in the doing guys. The data is there for you to look at.

Why not make available a processor for code only solutions for LOCUS or PM2? I realize that the aim in the beginning of Locus was to make it simple and affordable, but the unneccessary limitation of only being able to do cm level work with the Locus is problematic.

Shawn

Trying to reduce centimeter to sub-meter...I know what were talking about here, but on the big board, we'd be hung and strung...

Chris...The reason certain receivers get sub-meter reliably is because they were designed to use the CODE, not the carrier...The CODE is very reliable in getting sub-meter, but because of it's length and the resolution you can determine where the code starts and stops, cannot be used for centimeter..
The L1 receivers measure the phase carrier itself...
No big deal, just two different types of equipment...

Bob...I have gotten much better results using the CORS on baselines this long or longer...I'd expect 4-10cm on any occupation over two hours...and that's to one CORS..When you mention that you use 2 stations, how do you average the position, or do you let the software use the baselines to determine the solution? What type of loop closures do you see in the solution?


TM

I have $.10 to add (inflation). In the very near future, I will be making a similar decision on equipment for my fuure business. I also just completed a GPS class at ODU.

GPS is big bucks but with the right selection, will save $$ over time. There are a few requirements to be addressed. (1) Radio transmission distance - bigger output is better (within reason - allows longer distances from base station if needed. (2) L1 & L2 is best - allows better resolution of satellite information by mitigating ionisheric disruptions and increases accuracy and allows RTK which shortens collection times(3) powerful data collectors/computers at the collection location. Can get better information earlier and better point resolution. So for me - L1/L2, big radio (or cell phone), RTK with high powered data collector = survey grade =$$$. I am starting my savings plan now.

Ian