Determination of the National Highpoint of Botswana using GNSS

Reading time: 5 minute
...

📝 Original Info

  • Title: Determination of the National Highpoint of Botswana using GNSS
  • ArXiv ID: 2511.12332
  • Date: 2025-11-15
  • Authors: - 제공된 정보 없음 (논문에 저자 정보가 명시되지 않음)

📝 Abstract

Botswana has not previously been surveyed with sufficient accuracy to determine the highest peak in the country. Otse Hill and Monalanong Hill have been identified as the highest peaks, but there was uncertainty on which is highest. For this study, ground surveys were conducted on each of these peaks using an Abney level to identify the peak location and a GNSS unit to measure the elevation of each hill with sub-meter vertical accuracy. Monalanong Hill was measured to be 1.87m +/-0.02m taller than Otse hill, with an elevation of 1492.12m +/-0.01m (orthometric height, EGM2008 geoid, 95% confidence interval) above mean sea level (MSL). Otse Hill was measured to be 1490.25m +/-0.01m. Thus, Monalanong Hill is the highest hill in Botswana.

💡 Deep Analysis

📄 Full Content

The location of the highest peak in Botswana is geographically significant. The location of the highest point can be a tourist attraction and therefore uplift the economy of the local community. However, before this study, the elevation and location of the highest peak was not known with certainty. Previously, conventional survey methods and satellite-based measurements have been conducted.

Otse Hill was measured by a trigonometric ground survey in 1999 by the Botswana Department of Surveys and Mapping (DSM) (Botswana 1999). This measured Otse Hill 1491m, but Monalanong Hill (Table 1) was not measured. The purpose of the survey was to place a geodetic station. Hence, they did not observe the highest peak on the plateau, but a point suitable for geodetic control. In the current survey an Abney level was used to identify the high point on Otse hill.

In February 2000, the satellite-based Shuttle Radar Topography Mission (SRTM) collected elevation data at discrete points around the world, including in Botswana, using radar interferometry. The data was processed to create earth’s surface topography at a spatial resolution of 1 arcsecond (approximately 30m) as of September 2014.

According to Farr et.al and Smith 2023 et. al, elevations from SRTM have a vertical accuracy +/-16m. Elevations of locations between grid points of the SRTM Digital Elevation Model (DEM) are estimated. However, the error bounds of elevations of locations between measured points is unknown and can potentially be higher than +/-16m, especially for sharp peaks (Sandip 2013).

Digital elevation models based on SRTM data include Google Earth (Google Earth 2025), Google Terrain (Google Terrain 2025), Topographic-map (Yamazaki 2017), Floodmap (Floodmap 2025), and Gaia (Gaia 2025). All of these models identified the two highest peaks in the country as Otse Hill (location 25°0'27.6732" S, 25°25'56.0376" E) and Monalanong Hill (location 24°50'25.5732" S, 25°39'54.8244" E) (Fig 1). All other peaks in Botswana were lower and outside the error bounds of the elevation of these two peaks. The peaks have the same elevation within the error bounds of the measurements, and thus satellite-based measurements are not sufficient to distinguish which peak is highest (Table 1).

Thus, based on all existing measurements, it was unknown which of the two peaks is highest. Since the survey by DSM was not for determining the highest point and satellite-based methods only provide estimated elevations, it is imperative to carry out a ground survey to measure accurate elevations. GNSS offers the most efficient and accurate way to collect data. However, GNSS heights are ellipsoidal heights, which are simply geometric heights with no physical attributes. Hence the observed ellipsoidal heights will be converted to orthometric height using global geopotential models.

This study was based in South-eastern Botswana, where the two hills are located. 2). Processed GNSS data will yield horizontal and vertical position, with the vertical position being the ellipsoidal height. This height needs to be converted to an orthometric height, which will be done by using the relationship between reference ellipsoid and geoid as given in equation 1.

Where H is the orthometric height, h is the ellipsoidal height and N the geoidal height or geoid undulation (Fig 3). N is the height difference between the reference ellipsoid and geoid at a specific point on Earth. N can be derived from local geoid models, however where there are no local geoid models, global geopotential models can be used. In this case, N will be extracted from the geoid height calculator (UNAVACO) which uses EGM2008.

Elevations are reported as orthometric height using the EGM2008 geoid (Pavlis 2012), because this is the most accurate geoid for this area of Botswana. We converted ellipsoidal heights orthometric heights using the UNAVCO/NSF GAGE tool (UNAVCO 2025). All error bounds will be reported as 95% confidence intervals.

We processed data with PRIDE-PPPAR (Geng 2023) Results from TrimbleRTX static processing resulted in errors in elevation of +/-0.092m

for Otse Hill and +/-0.106m for Monalanong Hill. Monalanong Hill was found to be 2.07m +/-0.14m taller than Otse Hill.

We processed using PRIDE-PPPAR to first find kinematic solutions (Fig 5). This showed vertical errors for Monalanong Hill of +/-0.0137m and for Otse Hill +/-0.0105m. We next processed the data with PRIDE -PPP-AR to find a static solution for each location, which gave vertical errors of +/-0.00054m for Monalanong Hill and +/-0.00071m for Otse Hill. PRIDE PPP-AR solution converged faster when compared to CSRS-PPP. Static solutions generally have smaller errors than kinematic solutions, and final results will thus be reported as the static results. All methods found the ellipsoidal height of Monalanong Hill to be 1.368 m higher than Otse Hill (Table 3).

The geoidal height at Monalanong and Otse were calculated using the coordinates from above an

Reference

This content is AI-processed based on open access ArXiv data.

Start searching

Enter keywords to search articles

↑↓
ESC
⌘K Shortcut