Automated Data Collection (GPS and GIS) Part 2

Position Accuracy and DOP

Prior to the end of 1993, GPS had less than the full complement of 24 satellites operating. In earlier years, there were periods during the day when there were not enough satellites in view from a particular point on the ground to provide a position fix.

Now a data collector can almost always "see" enough satellites to get a position fix. But the quality (accuracy) of that fix is dependent on a number of factors, including

• the number of satellites in view, and

• their geometry, or arrangement, in the sky.

In general, the more satellites in view, the better the accuracy of the calculated position. The GeoExplorer receiver you are using, and many others, use the "best" set of four satellites to calculate a given point—where "best" is based on a DOP value. With more satellites in view, there are more combinations of four to be considered in the contest for "best." Data collection with fewer than five satellites in view is pretty iffy and should be avoided when possible. I’ll say more later on how you can know ahead of time when enough satellites are available and what the DOP values will be. It is almost always possible to see four (with 24 up and broadcasting properly—"healthy," as they say) and you might occasionally see more than twice that number.

The Geo3 is capable of "overdetermined position finding." This means it can use more than four satellites to determine a position. Since a set of four Can determine a position, a set of five could determine five positions—by considering satellites 1234, 1235, 1245, 1345, and 2345. The average of these positions would probably be better than any one alone. The statistical treatment of the use of more than four satellites is a complicated matter and will not be addressed further. Presumably overdetermined position-finding results in a more accurate result.

The concept of DOP involves the positions of the satellites in the sky at the time a given position on the ground is sought. To see why satellite geometry makes a difference, look at the two-dimensional case again shown by Figure 2—6. Suppose first there are two satellites ("a" and "b") that are being used to calculate a position "x": If we know the distances La and Lb exactly, we can exactly find the point "x."

But we don’t know La or Lb exactly, because of the error sources listed above. For illustration, suppose that we have an error distance, delta, that must be added and subtracted from each of La and Lb. That is, for each distance there is a range of uncertainty in the distance that amounts to two times delta. This is illustrated graphically by Figure 2—7.

Now suppose there are two satellites ("c" and "d" in Figure 2—8) which are being used to calculate a position "y." These satellites are further apart than "a" and "b," so that the angle their lines make at the receiver are more obtuse than the almost-90° angle from "a" and "b."

If we know the distances Lc and Ld exactly, we can exactly find the point "y." But again we do not know the lengths exactly—we use the same difference of two times delta (in Figure 2—9).

The shaded figure shows the range of positions the receiver might indicate for fix "y." As you can see, the distance between the true position and the position that could be reported by the receiver for the second case is considerably larger than for the first case.

My goal here is to illustrate that "satellite geometry" can make a big difference in the quality of the position you calculate. In actual GPS measurements, of course, it is a volume rather than an area that surrounds the point being sought, but the same general principles apply.

Satellite positions relative to "x."

Figure 2-6. Satellite positions relative to "x."

Area in which "x" might reside, given satellite positions "a" and "b." The shaded figure indicates the area that contains the actual location sought.

Figure 2-7. Area in which "x" might reside, given satellite positions "a" and "b." The shaded figure indicates the area that contains the actual location sought.


Satellite positions relative to "y."

Figure 2-8. Satellite positions relative to "y."

So, Actually, What is DOP?

DOP—sometimes referred to as GDOP (‘Geometric’ Dilution of Precision)—is a number which is a measure of the quality you might expect from a position measurement of the GPS system based solely on the geometric arrangement of the satellites and the receiver being used for the measurement.

Larger area of uncertainty, due to satellite positions "c" and "d."

Figure 2-9. Larger area of uncertainty, due to satellite positions "c" and "d."

If you think in terms of "geometric strength" you can pretend the electronic ranges from the satellites are cables and you want the strongest arrangement you can have to keep the receiver antenna from moving much when the cables stretch. Ideally you would like to have the four satellites at the vertices of a regular tetrahedron (a three-sided pyramid with all four surfaces equilateral triangles) and the receiver antenna at the center. This would provide the greatest geometric strength, but of course it is impossible for GPS, because three of the satellites would be below the horizon. The best alternative would be when one satellite is directly overhead and the three others are close to the horizon, spaced on a circle, separated from each other by 120°. Simply put, you get the best DOP when the satellites are spread out; the more satellites clump together, the worse the DOP value. In summary, DOP is a measure of the extent to which satellite geometry exacerbates the other errors that may occur in the measurement.

The overall DOP number is made up of several "sub-DOPs":

• HDOP (Horizontal DOP) is a combination of NDOP (North DOP) and EDOP (East DOP),

• VDOP is Vertical DOP,

• PDOP (Position DOP—generally considered the best single indicator of geometric strength) is a combination of HDOP and VDOP (actually the square root of the sum of the squares of HDOP and VDOP),

• TDOP is Time DOP, and

• GDOP ("Geometric" DOP) is a combination of PDOP and TDOP.

PDOP is the most important single DOP to consider. As you saw, on one screen you set the PDOP value; on another screen, during data collection, you got a report on the value of PDOP of the constellation the receiver was using. The PDOP mask says: "If the PDOP is too high, then don’t record any positions." The recommended PDOP mask value settings are: 1 to 4—great, 5 or 6—okay, 7 or 8—marginal, greater than 8— unacceptable.


In this session, using the same geographic location as last time, you will use the memory capacity of the receiver to store the readings in machine-readable (i.e., computer-readable) format. Later, you will also take data as you move the antenna along a path by walking, bicycling, or automobile. During both data collection sessions the data will be automatically collected into computer files.


Inside: Planning the GPS Data Collection Session

For the projects in this topic you will need your notebook, the receiver, and, optionally, the external antenna, if you have it. Set aside a section of the notebook to record information about the files you will collect. On a form, such as the one found at the end of this topic, you should plan to manually record:

• the date and time,

• a general description of the location or path that is the subject of your data recording,

• the starting time and the ending time of the data collection,

• the file name,

• the amount of memory available in the datalogger before data collection, and then, when finished collecting a file, the actual size of the file,

• the interval between collected data points (e.g., every 10 seconds, or every 50 meters),

• the values of PDOP,

• the number of satellites being tracked,

• the hours of charge remaining in the battery pack, and the amount of time it used during this session.

Setting up the Receiver/Datalogger2

Your goal here will be to ensure that the data you collect during your trip to the field will be worth something when you get back.

The receiver obtains data from the satellites and calculates positions at faster than one per second, but the datalogger is usually set by the user to record point fixes in the microcomputer’s memory less frequently. You have some control over how often point fixes are recorded. The datalogger also records the exact time each point fix is taken—a fact whose importance will become apparent later.

{_} Be sure that the battery pack is sufficiently charged to complete the session with an ample reserve. You may be able to get an indication of this by looking at how long the receiver has been used since the battery usage number has been reset: hold down the on-off button momentarily and note the number of hours of battery usage since the timer was last reset. Of course this number may be worthless for assessing the charge in the battery if it wasn’t reset when the battery was recharged, if a different battery was used by the last person taking data with the unit, or if the unit was powered by an auxiliary power supply as might be found in a car, boat, or airplane. In other words, don’t rely solely on the battery usage number when considering the charge left in the battery. When in doubt, recharge.

The amount of charge (in percentage form) in the internal battery of the Geo3 is displayed on the screen you get with Fn ~ OPTION ~ Status.

{_} Ensure that the receiver settings are proper. First check that the settings correspond to those below. Get to the screen "Main ~

Configuration ~ Rover Options" (As you recall, "Rover" means roving receiver, one that may be moved from place to place.) Scroll this menu until "Dynamics" appears.

{_} Set Dynamics: Land

The LAND/SEA/AIR choice relates to a number of factors affecting both display and internal operations of the receiver.

In the LAND mode the receiver is programmed to cope with "canopy" such as trees and heavy precipitation. It is expected that the antenna will be stationary or moving at relatively low velocity (e.g., automobile speed).

In the SEA mode, canopy in the form of precipitation is also expected, as is low overall velocity. But since the antenna might be mounted on a mast which could move rapidly from side to side, the receiver is programmed to expect such movement and disregard it.

In the AIR mode the receiver expects the antenna to be moving fairly quickly and that obstructions to the signal will be minimal. Also, in AIR mode, the display always shows directions in degrees based on magnetic north, rather than true north; pilots navigate on the basis of magnetic north, and runways are numerically designated based on their directions relative to magnetic north.

{__} Set Pos Mode: 3D

{__} Set Elev Mask: 15

{__} Set SNR Mask: 4

{__} Set PDOP Mask: 6

{_} Ignore the PDOP Switch (it is used for 2-D settings only)

{_} Set Antenna Ht: 0.00, or to the height you expect the antenna to be above the point whose 3-D position you are trying to record.

{__} Set Log DOPs: Off

{_} Set Velocity to Off. This option specifies whether the datalogger is to calculate and record the velocity of the receiver antenna. Velocity data use up memory and are particularly unuseful when the antenna is standing still.

{_} For the setting of "File Prefix" ask your instructor or leave it unchanged.

{_} Under the category of "Feature Logging," ignore all settings: "Points," "Line/Area," and "Min Posn." (The GeoExplorer has the capability of recording attributes of features (e.g., lampposts).

{_} Set the "Not in Feature Rate": The datalogger usually does not record all the points the receiver calculates. Which points are actually recorded? Basically, you can set it to record points under one of two circumstances:

• when a certain interval of time has elapsed since the last point was recorded, or

• when the receiver senses that the antenna has moved at least a certain distance from the last point recorded.

To experiment with this setting, press "CMD" when the highlight is over the number. On the screen that appears, you may, with the left-right arrow keys, select either of two fields.

The left field may be set to any integer value from 1 to 999, or to "All" or to "Off." (By continually pressing an up-down arrow key, you may scroll through a sequence of numbers; also, individual digits are addressable by use of the left- and right-arrow keys.) The right field may be set to "seconds," "meters," or "feet." Since in this exercise you will not be moving the antenna once datalogging starts, set the interval designation units to seconds. Set the number to three (3) and press "CMD."

To set the "not in feature rate" on the Geo3: Fn & OPTION ~ Setup ~ Configurations ~ Data ~ Log between features.

{_} Double-check that the "Not in Feature Rate" is set to three seconds.

{_} Under the category "High Accuracy" ignore all settings except "Recording." Make sure it says off.3

{_} Now you should have scrolled back to "Dynamics," which is where you came in. "Esc." "Esc."

{_} From the Main Menu, select "1. Data Capture."

You probably now have the following six options:

{_} Select "Open Rov. File" to open a rover file. The datalogger would begin recording data if the unit were outside and exposed to enough satellites that met the configuration requirements.

- Data Capture -

I.Operi Rov. File

2.0pen Base File

3. Review File

4.Delete File

5. Rename File


The data capture screen is transformed; it now looks something like this:



1 .File Status

2. Pause

3.Close File

4.Main Menu

The middle six characters hint at the date and time: the first two digits (mm) are the month, the next two (dd) the day, and the final two (hh) the hour in UTC time. (The year the file is collected is not reflected in the file name, but it is recorded, along with many other parameters, in the file itself.)

The final character ("i" for index) is one of the 36 character sequence "A, B, C,…Z, 0, 1…9." Prior to recording the first file of a given hour, the "i" character is set to "A" by the receiver at the beginning of the hour. It is "incremented" to the next character in the sequence each time a file is collected and starting during that hour. This allows several files (up to 36 with the same "u" character) to be collected during a single hour, each with a unique name. If a file is open as the hour changes the file name does notchange. Data collection continues under the same file name. A given file may contain data taken over a several-hour span.

{_} Check the file name displayed, to be certain it corresponds to the default discussed above, in terms of time and date. Record its name on a piece of scrap paper, since you will want to erase it shortly and you want to be sure you have the right one.

Before opening a file on the Geo3, set the datalogger so it beeps every time it takes a fix. This is the only way to know easily that it is collecting data. To set the "Beep volume" on: Fn & OPTION ~ Setup ~ Configurations ~ Other ~ Beep volume ~ On. Then CLOSE ~ CLOSE ~ CLOSE.

To open a file on the Geo3: Fn ~ OPTION ~ File. At this point you will see the possibility to "Create new file"; select that with the arrow keys if necessary. Before you press ENTER look below the two option boxes and find the file name. Confirm to yourself that it conforms to the description of file names given above in terms of date and time. Now press ENTER once only. You might think that you would press ENTER again, if you want to take data "Now" but that will tell the datalogger to start taking data for a Point, Line, or Area feature and you don’t want to do that.

Although it looks like nothing is happening, the receiver is feeding "Not in Feature" position data to the datalogger IF it sees enough satellites that meet the requirements. You will hear a beep from the unit when fixes are recorded. Of course, right now you are inside so no data will be collected.

{_} Note the number to the right of the file name. It is the number of data points collected. The number will be zero unless the antenna is outside and some data have been collected.

The menu items suggest that you may inquire as to the status of the file, suspend the data collection process, close the file, or return to the main menu.

{_} Scroll along the menu until "File Status" is highlighted.

Press "CMD." Scroll this menu until the name of the file appears at the top. The information portrayed should look something like this:

- File Status -

File: R112120A

Size 0.7519K

Free 188.82K

#Features 0

#Positions 0

So you have the file name, the number of fixes (points) in the file, and file size (in k, meaning kilobytes).5 Also on this screen the item "Free:" indicates the amount of memory space remaining in the datalogger. Depending on what else is stored in the memory, this number can be anywhere from about 200,000 characters (or "bytes") to none at all. When you want to know how much memory is left in the datalogger this is one way to see it.

Another designation is the "Size:" of the file currently being collected. A single 3-D fix, with no velocity or DOP information recorded, requires 20 to 25 bytes under the best circumstances. Under less than ideal conditions (e.g., interference with the satellite signals—by buildings and trees—which causes the receiver to switch constellations), each point may occupy an average of 100 bytes, more or less. Note that even though no "Positions" have been recorded in this file, it still uses up memory.

{_} How much "Free" memory remains in the datalogger?

About how many position fixes could you store in that amount of memory under the best of conditions?_About how many under poor conditions?_

{_} "Pause" the data recording process. "Resume" it.

{__} Close the file.

{_} Review the file: If there are several files in the memory you may scroll through them. They are numbered in the order that they were formed. The one you just "collected" should be the last. Check the name you wrote down on the scrap of paper. Several items of information are available through this file menu. For example, a file might look like:

{_} Move to the file you just opened—which will show zero positions. Delete the file. (Be careful to delete only the file you created.

Review: A010203A

Rover File

Log Rate: 10sec

Positions: 180

Features 0

Size: 7.055K

Duration: 1:05

It will be the one with the highest sequence number; also check that it was made today ("today" in Greenwich, that is).

{_} Turn off the receiver.

In the Field: Collecting Data

{_} Turn on the receiver.

{_} Return to the site of your first data collection effort (PROJECT 1B), where you wrote down position fixes on paper. Once you have considered the factors you previously learned about, regarding good data collection:

{_} Position the antenna where it was before,

{_} Wait until the receiver is locked onto enough satellites,

{_} Make the appropriate notes on the data collection parameter form found at the end of this topic, and

{_} Begin recording data: On the main menu select "1. Data Capture." Then select "1. Open Rov. File." Be very careful that you don’t open a base file. Very unfortunate things happen if you get this one wrong. ("Open Base File" is the option used when the receiver serves as a nonmoving station to collect data that may later be used to correct the data obtained by roving stations. Among other concerns, when a base file is collected, the receiver records data from all the satellites in view, not just the four at a time used for normal, 3-D fixes, and that uses up memory at a great rate.)

Begin collecting data with the Geo3 as described above. Be sure the Beep volume is on.

It has been suggested that about 180 fixes, if taken over a period of several minutes, are statistically sufficient to achieve a reasonable level of accuracy—given the quality of your receiver and the situation.

Recall that you set the datalogger to record a point every three seconds, so nine minutes should suffice for the data collection effort. If you have not collected the correct number of points in that length of time it is perhaps because, at times, the receiver lost lock on enough satellites. The reasons for this vary widely, but a principal one is that someone or something is getting in the way of a signal. The datalogger will record a fix only if DOP requirements, signal-strength requirements, number-of-satellite requirements, and so on, are met.

{_} While data are being recorded, take time to flip to some other screens; data recording will continue unimpeded. You can look at File Status and GPS Status. You may change such settings as datum, time display, units, and so on. Recording will continue (provided you keep your head out of the way of the satellite signals).

Important fact: It does not matter what display settings you use fordatum and units; the display has no effect on what is recorded in thedatalogger memory. Point fixes will be stored in latitude, longitude, andmeters above the reference ellipsoid. The lat-lon numbers will be basedon the WGS84 datum.

{_} Write in your logbook, on a copy of the form provided, the relevant information about this file.

{_} After nine minutes, verify that you have approximately 180 fixes. If not, figure out why. If you are well short of the target number of fixes, consider running the data collection part of the experiment again. Perhaps raising the PDOP to eight (8) will help, although it may slightly degrade the quality of the data you get.

{_} Close the file. Review it. Finish writing in your log.

Assuming everything went according to plan, turn off the receiver.

{_} At this point you may either go on to Project 2C, in which you upload the data you have collected into a PC, or you may collect additional data in Project 2B.

Next post:

Previous post: