GPS and surveying...

[rusty0101]

I fully realize that my handheld GPS, while a very capable instrument for general use, is like a hand adze in comparison to a surface planer when compared to a surveyors sub centimeter GPS (I presume that this is also a dopler based gps). In other words both a planer and a hand adze will give you a surface that can be used, but the surface planer will give you such a surface much more accuratly.

 

My question is how is a surveyors gps used? More specifically my experience with a gps is that if you are going to use it for course evaluation, you turn it on when you start, give it some time to settle down with good readings, then leave it on till you get where you want to be. I would presume that a similar situation exists for a high precision GPS.

 

In a couple of other threads I read that surveyors start from a known benchmark with their GPS and work off of those readings. My suspicion is that a surveyor goes to that benchmark, lets the GPS aquire a stable reading, then the surveyor enters the absolute coordinates for that benchmark. From that point on, the GPS remains on, and as much as possible with open sky above it, while survey work is done. In situations where it is possible that the GPS may loose sattelite data, or the GPS reports that something seems odd, the surveyor may return to a benchmark, or local temporary benchmark to re-calibrate the equipment..

 

Am I far off?

 

-Rusty

[+GPS_DUDE]

Hi Rusty,

 

I use an expensive GPS unit for surveying in the forest industry. After a days work in the field, collecting data, which includes GPSing roads, creeks, crossings, and cutblock boundaries, the data logger is downloaded to a PC. A GPS base station is used as well which collects postions on the satalites. With post processing software, the data is combined and a differential corection is done. This gives a sub meter acuracy. From here the data is entered into a (GIS) Geographial Information System as a layer which produces digital maps. Another option is working in real time, which has the GPS unit linked to the base station via handheld radio.

[Prof. Y. Lupardi]

quote:

---

Originally posted by rusty0101:

[...]My question is how is a surveyors gps used? [...]In a couple of other threads I read that surveyors start from a known benchmark with their GPS and work off of those readings. My suspicion is [...]

-Rusty

---

 

You do not have to be a surveyor or have super-expensive machines if you have time on your hands. I have experimented with a GPS12 and managed to get the position of a corner of my roof within 20 cm. I also tested benchmarks of known position to verify this. More you can read on my webpage:

http://www.xs4all.nl/~atarist/geo/gps_accuracy.htm

[+AzBat]

Hey Rusty,

 

I run a GPS crew for my dad's surveying business. Oddly enough I hadn't done any surveying for about 10 years prior to this past January when I started to work for him. It was my first experience with handheld GPS and GeoCaching that lured me to the job my dad was offering.

 

Since January I've learned so much about using GPS for land surveying. I suspect there are professional surveyors lurking around here that have used it longer, but I figured I could give you a simplified explanation coming from somebody that had used a handheld GPS first.

 

Anyway, GPS in surveying is used in many different ways. There are boundary surveyors, like my dad, and construction surveyors. Usually they require the sub-centimeter accuracies you were talking about. There are also GIS people like GPS_DUDE, who may only need sub-meter accuracies. For boundary surveying sub-centimeter is an absolute must and thus we use the more expensive GPS receivers out there. Of those there are 2 different kind of receivers: single frequency and dual frequency. Single frequency can only use the L1 signal and therefore is used mainly in static GPS collection(I'll explain more later) and is much cheaper. Dual frequency uses both the L1 and L2 signal, so it makes it more suited for doing kinematic surveying, but we've used it a lot on static GPS surveying since baselines can pretty much go hundreds of miles.

 

Back to static GPS collection. We use this in surveying mainly to establish control points for our surveys or our county-wide control network. With any GPS surveying you're going to need at least 2 GPS receivers setup on 2 different points at the same time. They need common time between them, which I'll explain more later. Anyway, Static means we setup GPS receivers on tripods with an antenna over 2 different points. We occupy both points for a period of time(10 or more minutes) to let them collect data from the satellites. The precision of each point is dependent on the amount time we occupy the point, the number of satellites, their location in the sky and any obstructions at the location. I've seen times where we were at a location that had no obstructions, had 8 or more satellites and we are able to get an accurate position(about half an inch or less) in 20 minutes. But I've also seen us on a side of a hill with lots of trees, had only 4-6 satellites and we stayed on the point for an hour and a half to get a position that was only accurate within a 2 or more inches. In surveying, you need at least 4 satellites to get a horizontal position and 5 satellites for a horizontal and vertical position. We always try to use more than 5 satellites.

 

After you've collected the data on the receivers you can download it to a computer. This will allow you to use the data in a program that will allow you to do post-processing. By setting up 2 different GPS receivers on 2 different points at the same time we are able to create a vector between them. The software will do calculations on the collected data from both receivers to get a an accurate distance for the vector between them. If one of those receivers was setup on a known coordinate, then we are able to get a fairly accurate position of the second point. However, it all depends on the conditions of the site. If we had a third receiver setup on another point running at the same time as the other 2 receivers then when we bring the data into the software we would have 3 vectors making up a triangle. This would give a much more accurate position of all 3 points since now they have a second vector coming into each point, sort of a check. So you can see that creating triangles is the best way to get the most accurate positions. Also, the more vectors coming into a point the more accurate you're able to calculate a position.

 

One bad thing about static GPS surveying is it's only good to about 7 or so miles if you're using L1 receivers. However, the biggest problem is that it takes a lot of time to get good positions. That's where kinematic or real-time kinematic(RTK) is much more suited since you can occupy points with substantially less time. The main difference with RTK GPS surveying is that now the GPS receivers come with radio modems. You have a GPS receiver called a base that contains a radio modem that sends differential corrections to another GPS receiver called a rover. The radio on the base usually has much more watts of power in order for it to able to transmit its corrections over 5 or 6 miles. The rover can run with a much smaller powered radio. Our base runs at 35 watts and our rover runs at 2 watts. Anyway, with RTK all the work is done in the field in real-time. There's no need to take collected GPS data to the office to post process. In order to do this you'll need some kind of data-collector or hand-held computer to give you the positions. The receivers themselves don't do any calculations. The base receiver just sets there transmitting data to the rover. The rover receives the data and the data collector will take that data and the data the rover itself receives to calculate the positions. However, before you can get started shooting points you'll need to setup the base and do what is called a localization. When you setup your base receiver it can be on a known or unknown position, it doesn't matter. You tell the receiver to get an autonomous position in geodetic coordinates and set the occupied point to that position. Next you can take the rover and shoot 1 or more known points. This tells the software in the data-collector how to calculate grid system for the job area. If you're not going to use a grid coordinate system like State Plane, then you can probably localize on one point for a reference frame and you'll get true ground distances and bearings, but the more points use to localize the better. If you want to use State Plane coordinates or another grid system, then it's best to use a box of known control points and work inside that box. That's what we usually do. Localize on 4 points of a box usually covering a mile square around our job site. When you shoot a point(in your localization or others), it will collect data every second. It will compare this data with the corrections it receives from the base and the satellites it uses. Depending on the conditions at that site, it might take 10 seconds to 10 minutes or more to get a good solution. During this time the data-collector will give you information about the number of the satellites, where those satellites are located, quality of those satellites, the accuracy of the solution, etc. Once you think the accuracy is to your requirements, then you can store the point and go to the next point. Most of the time when we do RTK work, we'll set on a point about 30-60 seconds. So you can see that with this method you can get much more work done than with static GPS surveying.

 

There's a lot more to GPS surveying than what I talked about above like multi-path, mask angles, recording intervals, DOP(dilution of precision), RMS(root mean square), SNR(signal to noise ratio), CORS(continuously operating reference station) etc. If you have any questions, or want me to get more detailed then let me know.

 

Tommy McClain

[+Kewaneh & Shark]

The GPS equipment that is used by surveyors works on the same principles as your hand held - triangulation by satellite signals. The biggest difference is in the equipment, and the setup and use of that equipment.

 

Generally, a small handheld GPSr can be give a fairly accurate reading when the conditions are right. The conditions aren't always right though, and most handhelds have an inherent plus/minus 10 meter (33-35 feet) error factor in them. WAAS enabled handhelds can be down to 5 meters error.

 

One of the biggest equipment differences in surveying GPS units would have to be the use of a base station with a roving anteanna - basically two parts. The base station is a non-moving station that receives constant communication from the available satellites. It may or may not be positioned over a known point, such as a benchmark. As the base station receives more data from the satellites, its position gets a more refined (a more stable reading), - a process called averaging.

 

The rover also receives data from the satellites and works very similarly to the handhelds that we use in geocaching & benchmarking, with the exception the it is also in communication with the base station, and knows its position relative to that base station. As the position of the base station gets refined by the averaging process, the position of the rover gets more refined also. The data obtained by the rover is stored in a seperate data collector which may or may not be attached to the rover.

 

As GPS_DUDE mentioned, the data obtained by a surveyor is then downloaded into a computer for post-processing. It cannot be taken at face value in its raw-data form. The post processing adds any necessary correction factors and makes adjustments for the satellite geometry and period of time of the GPS session. (Plus a multitude of other things that can only be understood with a Ph.D. and witchcraft.)

 

Some of the other differences in survey-grade GPS equipment would be the size of the equipment, in both dimensions and power use. Dimension-wise, the rover can be 6-8 feet tall and weigh a few pounds. The base stations can be taller. I've seen base station anteannas placed as high as 15-16 feet. The battery on the rover is fairly specialized (each maker has their own) and compact, but the base stations, particularly older models, are commonly attached to car batteries. Also, survey systems use much higher grade anteannas, which have had their phase centers (the point that is being measured to) calculated. Many also have the ability to detect multi-path signals - a reflected signal that will give an erroneous reading - and they have the ability to disregard those signals.

 

It needs to be remembered though that a high precision GPS does not change the aspect of surveying. Proper surveying proceedures are still necessary in order to achieve the needed accuracy, such as returning to a benchmark. GPS is just another tool used by surveyors.

 

Surveyors do use handhelds, too. They are the best for getting close to a site location.

 

Keep on Caching!

- Kewaneh

[Kerry.]

One thing with surveyors (as Kewaneh alluded to) is not the fact of simply pushing buttons but "knowing" why one is pushing buttons AND then knowing what to do with the data from that button pushing.

 

Really there's quite a range of different types of GPS surveys based on accuracy and requirements specific to the type of survey and the requirements.

 

Real-time cm type accuracy has it's purpose but is basically restricted by distance as a function of radio transmission and ambiquity resolution.

 

With static it comes down to time and basically the longer the line the longer the time required. Where as with real-time stuff (RTK etc) receivers are running generally at 1 second epochs where as static might be recording every 30 second epoch but maybe for hours, days, weeks etc but distances in the thousands of km is possible to less then cm accuracy.

 

Machine control systems can be set to transmit 10 updates a second and some can actually do that (at a cost $$) where as lower cost units can manage 10 outputs a second but only based on a 1 second position update.

 

The systems are there to meet the needs and the $$'s basically follow.

 

Real-time systems such as the automatic Aircraft carrier systems run at 20 Hz (20 position updates per second).

 

The requirements are to land a carrier jet (like a FA-18) on a carrier at sea under auto pilot. This basically requires the hook to "fit" through a 3 foot square window 14 feet above the end of the carrier deck so as to meet the arrestor wire.

 

There are multi (as many as 6) movement dynamics based on the carrier doing X number of knots with the similar addition of the aircraft dynamics (roll, pitch, yaw, speed, drift etc etc) all on an angled run way at a closing speed that jets need to fly at.

 

The dynamics is mind blowing and it's one cool pilot putting their faith in a GPS landing system at that level. But there is a system and it's been tested under sea conditions.

 

That's off the track of conventional surveying but there's still a long way with GPS yet even better than cm or mm survey requirements.

 

Cheers, Kerry.

 

I never get lost everybody keeps telling me where to go

[+canadazuuk]

Great thread. I've been wondering about surveying and GPS ever since a multitude of caches put out by TEAM KFWB GPS showed up.

 

canadazuuk

[Z15]

Here's my 2 cents

 

I was the senior Survey tech for 31.5 yrs.

 

My equipment included 3 Leica GPS 500 receivers (the best), 1 base station and 2 RTK rovers, a 1999 4X4 Suburban and a 2001 Dodge 3/4 ton 4X4 ext cab pickup. I also had access to 6 other Leica receivers from other crews. We also have a network of CORS Stations MICHIGAN CORS all over the state, as well as NGS and Coast Gaurd CORS all over the Great Lakes.

 

What we did was install our own survey markers (driven rods-Bernsten) or used NGS/USGS/USCE/NOS points that were clear to the sky. My boss (PS) had a goal of setting XY on all the NGS bench marks were could find and use. Would set the 3 receivers on the control points to base our survey on, no more than 2 miles apart and often only 1 mile. Run them for 5.5 hrs on 2 separate days, once in the AM and once in the PM. At the end of every day, the data was downloaded into a Dell Latitude and converted to RINEX data, then uploaded to NGS OPUS for final processing. We would then mean the sessions and do the field RTK work. Usually took a week to set the main control depending on the project size. Once the precise orbits were available we would once again run the OPUS doe the final time.

 

Once we had the control points and while we were waiting for precise orbits (14 days latency), we would RTK (Real Time Kinematic) in our intermediate control, e.g. highway alignment points, plat corners, sections corners, mapping points. aerial control targets, etc. This usually involved driving along in the truck with the antenna out the window so as not to loss lock on the SV's. These points were tied for 2 seperate control points and averaged to predetermined specs. It took a minumun of 5 SV's to solve for position, x+y+Z+Time+Check. This was all done by the receiver, often in less than 1 second. You could learn to do this is 15 minutes, even the PS could do it. This data was then uploaded to the laptop, processed with Leica SKIPRO software. Any points that did not average, we would hit again the nect day.

 

Once all the field work was done, we would reduce the state plane coords to a project datum for the consultants and constcution forces to use.

 

This was often just a 1 man operation except for the summer, we used COOP students from MTU and Ferris to help out the crew. On one project last winter I ran 6 recievers at once using all NGS bench marks. I put on about 200 miles a day. Would set up the stations in the AM, program the receivers to turn on at noon and turn off a 5:30 pm. Area covered was about 20 miles and I just keep driving around checking on the equipment. It was along busy US-2.

 

That was fun work (GPS) but living out of motels and working out of trucks for 31+ yrs got to me and I took early retirement in Nov. I was only 1 tech but had 3 professional surveyors to deal with, each one had a different view of how things had to be done.

 

Image of our receivers set on a Order A station 2001 L 61 when we reobserved all the HARN in Michigan and postioned out CORS. It was not this point but I was sitting out in the boonies on a HARN point on 911 and just happened to turn on the truck radio at around 3pm.

 

 

[This message was edited by elcamino on December 05, 2002 at 05:34 AM.]

[gurubob]

Leica is the best... Have you ever heard of a little navigation company out of Sunny Vale California? Trimble is in my opinion the finest GPS survey equipment maker to walk the face of the earth.

In short... a minimum of 5 satellite vehicles(SAVS) transmit a series of coded frequencies L1, L2 & P (most common in private and military applications). A receiver like the car radio reads these signals and establishes its position. Static Surveys can require observations up to 2 hours. Faststatic does the same thing in about 7 minutes. since selective availibility was turned off even hand held consumer models can determine position to very accurate levels. RTK lets this assumed point act as a transit for radial collection and staking. A rover(second receiver) reads the SAVS signals and a radio link between the two receivers allows them to "talk" to each other establishing a relative position. The rest is math.

[Kerry.]

quote:

---

Originally posted by gurubob:

Leica is the best... Have you ever heard of a little navigation company out of Sunny Vale California? Trimble is in my opinion the finest GPS survey equipment maker to walk the face of the earth.

/quote]

 

I wouldn't call that an entirely a subjective statement that one.

 

Static Surveys can require observations up to 2 hours. Faststatic does the same thing in about 7 minutes.

---

 

Trimble's Faststatic is simply a trimble name, it not all that different to what other manufacturers implement at all, it works, as does all the other different types as Rapid static, Classic static of static. They all have a purpose.

 

quote:

---

since selective availibility was turned off even hand held consumer models can determine position to very accurate levels.

---

 

No, I wouldn't call handheld consumer models "very" accurate, not in the overall context anyway.

 

In general it is very close (actually already here) to the time where different manufacturers have had to rationalize there approach and thinking and realize that users of high end receivers want the capability of using whatever receiver for whatever purpose without having to be stuck in the "manufacturers" straight and narrow rut.

 

No longer is it acceptable to be only able to use brand X with brand X especially in a real-time situation as large scale projects have a range of contractors and there's no reason why they have to conform with brand X.

 

Cheers, Kerry.

 

I never get lost everybody keeps telling me where to go

[Z15]

http://www.ellisv.com/