The GPS Logger (version 3.2.3 at the time of writing) is is available for Android smartphones. It can be downloaded and installed from the Google Playstore.

Once installed, using the app is quite easy. Move out to an open area where you have visibility of the sky, enable the smartphone GPS (location) and start the GPS Logger app. Figure 1(a) shows the opening screen of GPS Logger 3.2.3 displaying a GPX fix after a few seconds of switching ON (the first fix might however take longer so be patient!).

Once you have a stable fix and a stable estimate of accuracy, you can start collecting waypoints and tracks. Referring to Figure 1(a), to record a waypoint or a placemark, tap on the Track Tab (callout 1) and press the Add Annotation button (callout 2) at the bottom of the screen. Give your waypoint point a name. You can then walk to another waypoint of interest and repeat the process i.e. press the Add Annotation button to record and then give the second waypoint a name. In this manner, you can collect as many waypoints as you want. Once you have done collecting your waypoints, press the Stop button (callout 3) to conclude and give this collection of waypoints a name under which these waypoints or placemarks will be saved in the tracklist. To see the saved waypoint file, press the tracklist tab.

To record a track, press the Record button (callout 4). Walk or drive around to complete your track and to stop recording, press the Stop button (callout 3) and save the track with a name. During the recording of a track, you can pause the recording by hitting the pause button (the record button becomes the pause button while recording) and annotate key landmarks on the track by pressing the Annotation button and giving that landmark a name. You can resume recording the track by hitting the record button again. When you are recording a track, you can, if you wish, press the lock button which prevents accidental pressing of the stop or pause buttons while recording. When the lock button is engaged, no buttons on the bottom bar work.

Press the Tracklist tab (callout 6) on top right of the screen to see all your saved waypoints and tracks, indicated by different icons (Figure 1(b). Notice in Figure 1(b) that the track icon indicates that track was completed on foot (walking figure).You can change these icons via the edit. Tapping on any item listed (Figure 2(a) — callout 1) in the tracklist opens up options to view , edit , share , export or delete (Figure 2(a) — callout 2). Try seeing a saved placemark or track on Google Earth if installed using the icon. Try exporting one as a KML or GPX file by tapping on the export icon. Tap the three dots to access export settings (Figure 2(a) — callout 7) from where you can choose export formats as seen in (Figure 2(b). To know more about using the app, view the tutorial videos mentioned below or refer to the getting GPS Logger get started guide.

These waypoints and tracks can be exported in GPX, KML and text or Comma Separated Values (CSV) formats. There is absolutely no limitation on the number of waypoints or tracks that can be recorded or exported.

In summary,the app is very easy to use, doesn’t track you and does not share any data with any third party. It starts almost immediately even on low end smartphones and has a remarkably clean interface that focuses on the essentials.

As is the case with most smartphones which have assisted GPS, you can get a much faster fix if prior to moving out in the field, you enable your smartphone GPS when you have internet access and get a position fix. This quickly updates the almanac information (which remains useable for a few months) via the internet and is particularly time saving if you haven’t used positioning applications for a long time or have moved to a new location a few hundred kilometers away from where you last used your smartphone GPS. Fundamentally, however, the GPS logger app does not need Internet connectivity and can work without any data assistance in the field.

Table shown in Figure 4 presents the horizontal accuracy estimates between Garmin GPS receiver (reference value) and the GPS Logger App (test value). The RMSE for the 20 sample points was computed to be 1.63 meters and the US National Standard for Spatial Data Accuracy Statistic (NSSDA - 95% confidence level) was computed to be 2.82 meters.

What the above results indicate in summary is that 19 out of 20 points collected using the GPS logger are likely to fall within about 3 meters in terms of horizontal accuracy from the readings collected using the recreation grade GPS receiver used in this experiment. Only 1 point out of the 20 points is likely to fall farther than 3 meters from the location reported by the recreation grade GPS receiver used in the study.

Three tracks in the same area were also collected using both Garmin GPSMAP 78s (blue line) and the GPS Logger App (red line) to assess visual correspondence (Figure 5). As can be seen, the tracks from both sources, by and large, match up pretty good with occasional divergence of 3-5 meters at certain places.

With reference to comparison of vertical issue between the Garmin GPSMAP 78s and the GPS Logger, it is instructive that we first dwell into a brief discussion on vertical datums. While this discussion might appear theoretically knotty for some, it is necessary to understand what the GPS Logger app is doing. From a broad working perspective, the concept of vertical datum and heights can be conceptualized with the help of Figure 6. The Mean Sea Level or MSL is defined as the zero elevation for a local or regional area (NGA, 2014) and elevations are commonly represented above the MSL. The GPS models the earth as an ellipsoid, which is a geometrical approximation of the size & shape of the earth. However, the earth while resembling an ellipsoid (squished at poles and bulging at the equator), is however not uniform and has mountains and trenches popping above or going below the uniform ellipsoidal surface (labelled topography in Figure 6). When a GPS receiver reports elevation at any point on the earth’s topography, it is usually the height from the surface of the ellipsoid to that point – denoted by h and called ellipsoidal height. The surface of the ellipsoid is not the Mean Sea Level (MSL). The Geoid instead, technically referred to as an equipotential surface defined by earth’s gravity, approximates closely the Mean Sea Level. The geoid in simple terms creates a mathematical model for the entire globe virtually extending the coastline (MSL) inland using gravity values. Witold (2003) and NGA (2014) provide an excellent explanation of the geoid concept in the following words:

“It turns out that MSL is a close approximation to another surface, defined by gravity, called the geoid, which is the true zero surface for measuring elevations. Because we cannot directly see the geoid surface, we cannot actually measure the heights above or below the geoid surface. We must infer where this surface is by making gravity measurements and by modeling it mathematically. For practical purposes, we assume that at the coastline the geoid and the MSL surfaces are essentially the same.” NGA (2014)

Image Credit Oklahoma University

So referring back to Figure 6, if we want to report the height above the Mean Sea Level or MSL i.e. H also known as orthometric height, we need to calculate N – the geoid height or the difference between the geoid and ellipsoid at a given coordinate on the surface of earth and subtract it from the ellipsoidal height.

The geoid height at any coordinate can be calculated using geoid models like EGM96 (Earth Gravitational Model 1996) which has been developed by the US National Imagery and Mapping Agency (NIMA), the NASA Goddard Space Flight Center (GSFC), and the Ohio State University. A good explanation of the geoid and MSL is available at this NOAA website.

With the above background, let us try to breakdown how the GPS Logger calculates orthometric height from ellipsoidal height after the EGM96 correction is enabled (Figure 7).

Figure 7 a shows the GPS Logger altitude value before EGM96 correction i.e. ellipsoidal height, Figure 7b shows GPS Logger altitude value with EGM96 correction switched on i.e. Orthometric height and Figure 7c, Garmin GPSMAP 78s altitude value at the same spot. Notice prior to the EGM96 correction, the GPS Logger altitude value is 487 meters and with the correction, it is 544 meters – a difference of 57 meters!

How did the GPS Logger arrive at 544 meters? For the purpose of illustration, I independently tested this correction using an online Geoid Calculator sponsored by NASA and US National Science Foundation. Figure 8 shows this calculation for coordinates and elevation (487 m) given in Figure 7a. Figure 9 shows the computed results. Figure 10 shows geoid height calculation for the same coordinates (IIFM Bhopal campus) using another EGM96 geoid calculator at the NGA site.

http://earth-info.nga.mil/nga-bin/gandg-bin/intpt.cgi

The geoid height computed by the UNAVCO tool is the same as the geoid height shown by the NGA (US Military EGM tool). As can be inferred from figures Figure 7b & Figure 9, that the orthometric height calculated by the UNAVCO tool is identical to the orthometric height calculated by the GPS Logger app. As is evident, the EGM96 geoid height for this coordinate is -57.091 meters, ellipsoidal height is 487 meters and so the orthometric height is 487 – (-57.091) = 544.091 or 544 meters as calculated by GPS Logger. This matches exactly with the elevation shown by the Garmin GPSMAP 78s (Figure 7c) and very close to the Google Earth elevation (MSL) at the same location as shown in Figure 11.

This discussion on altitude is important for another reason. When we try to compare orthometric height given by GPS Logger (Figure 6b), we can compare it only with orthometric height from the GPS receiver and not the ellipsoidal height. Many GPS receivers however display ellipsoidal height rather than the orthometric height. However, in recent times, major GPS receiver manufacturers, after determining the default WGS84 ellipsoidal height, correct the same performing behind-the scenes calculations using geoid models and then finally displaying orthometric heights. As this experiment was primarily aimed at understanding how the GPS Logger android app performs in comparison to a recreation grade receiver, in terms of altitude comparison, it was important to know if the Garmin GPSMAP 78s was indeed reporting orthometric heights as figure 6b and 6c seem to suggest. The user manual of the GPSMAP 78s did not categorically mention anything on this in particular and searches on the Garmin website were not useful even though the glossary of terms at the Garmin website refers to elevation as distance above or below mean sea level The author contacted Garmin for clarity on the above but could not get a response until the writing of this document.

To find out if the Garmin receiver was indeed computing and using geoid height, raw NMEA sentences were captured from the Garmin GPSMAP 78s through the com port and it was observed that the $CPGGA sentence (the sentence containing the GPS fix data) was indeed calculating the geoid height and therefore reporting orthometric height or height at MSL. Figure 12 shows a sample waypoint capture of the NMEA sentence stream. The highlights show the altitude for a waypoint in meters above mean sea level as 560.9 m and the height of the geoid (mean sea level) above WGS84 as -59.6 m as per NMEA $CPGGA sentence specifications.

The table shown in Figure 13 presents the elevation accuracy figures between the GPS Logger with the (altitude correction — EGM96 Correction — switched ON) and the Garmin GPSMap 78s. The RMSE was computed to be 3.54 meters and NSSDA Statistic (95% confidence level) to be 6.94 meters. The mean elevation difference (Column E) was -1.39 meters (absolute value 1.39 meters).

It must be mentioned here that very accurate height measurements (other than that obtained from GPS) of the 20 random points that were tested were not available. This was also not felt necessary as the objective here was primarily to compare the performance of the GPS Logger app with a standard recreational grade GPS receiver (rather than make an absolute assessment of which device was more accurate).

In summary, height measurements reported by GPS Logger, after the EGM96 correction was enabled, were usually quite comparable to those reported by the GARMIN GPS 78s provided there was no major obstruction of the sky.

These preliminary results seem to suggest that the GPS logger performed quite well and can be used like a recreational grade GPS receiver to collect georeferenced data for many applications that do not require highly precise positioning accuracy. The orthometric height from the GPS logger showed a mean elevation variation of 1.39 meters (compared to Garmin GPSMAP 78s) after switching on the altitude correction model option (EGM96 Correction available from settings). At some points however, the altitude was off by 5-8 meters when compared to the altitude reading at the same location from the Garmin GPSMAP 78s. Given the altitude NSSDA statistic, altitude readings reported by GPS Logger after switching ON the altitude correction model might be useful for many applications. However, careful assessment is advised to evaluate acceptability of altitude accuracy for applications under consideration. It is recommended that users can consider keeping the EGM96 correction option ON to directly get orthometric heights.

I would also like to mention that after enabling your smartphone GPS and the GPS logger app, it is advisable to maintain a stationary position for about 4-5 minutes for the signal to stabilize before proceeding to collect waypoints and tracks. In fact, it might be quite desirable to do a small mock exercise with a few sample waypoints giving the app time to stabilize the fix and the user to get used to the app in field conditions before starting the actual survey.

It is worthwhile to caution however that it will be prudent to treat the above-mentioned comparative estimates of horizontal and vertical accuracy as indicative rather than conclusive and try out the app in an exploratory exercise (and preview results on Google Earth if possible) to see if it meets accuracy criteria for a particular study. It is also important to be mindful that accuracy might vary in other terrain configurations like denser vegetation etc. or in situations where there is obstruction in the line of sight.