Wisconsin State Highpoint Survey

Wisconsin State Highpoint Survey

Oct 26, 2024

Eric Gilbertson

Summary of Results (NAVD88 datum)

Timms Hill 1951.9 ft
Pearson Hill 1950.8 ft

Timms Hill is highest point in Wisconsin, Pearson Hill is second highest point.

Results published in Progress in Physical Geography:

Gilbertson, E., Hensley, R., Kirmse, A., Bretherton, K., Stanchak, K., “LiDAR Accuracy on North American Mountain Summits,” Progress in Physical Geography: Earth and Environment, 2025 Link: https://doi.org/10.1177/03091333251401361 (Link to free version: https://arxiv.org/abs/2511.12341 )

Introduction

Timms Hill and Pearson Hill are two peaks in north central Wisconsin near the town of Ogema. They are approximately 1 mile apart and are the highest two hills in the state. Timms Hill had a fire tower constructed on its summit in 1932, and was first surveyed by USGS in 1934 [1] using third order leveling. It was later surveyed in 1974 using photogrammetry [2], and in 2004 [3] and 2012 [4] ground surveys.

In 1982 a lookout tower was constructed in the summit area next to the 1932 fire tower, and some ground leveling occurred. This may have affected the summit elevation.

Pearson Hill is about 1 mile east of Timms Hill and contains a wooded summit on private land. It was first surveyed in 1962 by USGS [1] using third order leveling, and later in 1974 via photogrammetry [2].

The 1934 and 1962 ground surveys showed Timms Hill about 1.5 taller than Pearson Hill, but the 1974 photogrammetry showed them the same elevation to the nearest foot.

The USGS monuments on each summit are much taller than normal – the concrete post holding the benchmark on Pearson Hill sticks up 1.2 ft above the surrounding ground, and the concrete post on Timms Hill sticks up 3ft above the ground. The reported elevations for each peak by USGS are so close that the heights of these concrete posts are significant. Generally, surveyors report the height of a benchmark when reporting results, unless explicitly specified otherwise in field notes. Benchmarks give repeatable locations so future surveys can measure the exact same location and compare results.

If the reported heights for each peak were in fact the benchmark heights instead of the ground heights, that would make the Timms Hill measurement artificially 3.0ft too high, and the Pearson Hill measurement 1.2ft too high. That might be enough of a change to affect which peak is the true highpoint of Wisconsin.

To add to the confusion, the 1934 benchmark notes for Timms Hill state the concrete post projects 1.4ft above the ground, and the 2004 survey field notes confirm this, but it currently sticks up 3.0ft above the ground.

Lidar data from 2018 [5][6] showed both peaks very similar in height, with Timms Hill about 1.5 taller than Pearson Hill. Lidar has a stated vertical accuracy of +/- 4in (+/- 0.33ft) for measured points [7].  However, Lidar can be subject to errors of mis-classifying dense brush as ground, or measuring artificial concrete structures as natural ground. I’ve found through my own research that LiDAR vertical accuracy for summit elevations on below-treeline peaks is +/-4.8ft (95% confidence interval errors) [13]. Timms hill has both dense brush and concrete foundations near the summit. So it is possible Lidar was giving an over-estimate for Timms Hill. Pearson Hill does not have dense brush or large artificial concrete structures near the summit, so Lidar was likely more accurate on Pearson Hill.

Given the confusion about which point the USGS surveys were measuring, and the potential error in Lidar measurements, I wanted to take ground measurements of each peak to find out for certain which one is taller.

I have surveying equipment capable of measuring elevations more accurately than Lidar, and I can measure the exact summit of a peak. I’ve surveyed dozens of peaks around the world over the years, and I can generally get elevations to the nearest +/-0.1 ft or better.

Methodology

I planned to take measurements with two survey-grade differential GPS units capable of vertical accuracies 0.1ft or better. A differential GPS unit is much more accurate than a standard handheld GPS or a phone, which can have vertical errors +/-20 ft or more. A differential GPS has access to many more satellites and uses an external antenna to help with multipath errors. It also collects enough data that results can be compared to base stations around the country to correct for atmospheric distortions. Data can also be post-processed by other sophisticated methods like PPP (precise point positioning).

I planned to take a 4-hour measurement at each location. This is twice the standard length of time for surveying benchmarks with differential GPS units.

From photographs and trip reports I knew Pearson Hill was forested with overhead tree cover. This is not great for GPS measurements since it can add error. The error can be reduced if the leaves are off the overhead branches. The Timms Hill monument is directly underneath the 1932 fire tower, with trees nearby. I suspected the fire tower structure would likely interfere with GPS measurements.

I determined the optimal time of year to survey these peaks was late October. Historically in northern Wisconsin all the leaves have fallen off the trees in this area by late October, but snow doesn’t typically start until November. I didn’t want to survey with snow on the ground because this would complicate finding the highest point on each summit.

For equipment I planned to use a Promark 220 dGPS and a Trimble DA2 dGPS. Using two different models of dGPS would help increase confidence in the final results. I would use 2.0m antenna rods to get the antennas far off the ground to increase accuracy.

Previous Measurements

I first conducted background research to fully understand all existing measurements of the peaks. This included researching all previous USGS measurements and conducting detailed analysis of existing Lidar measurements.

USGS

All USGS measurements for each peak are shown in Table 1, with elevations converted to NAVD88 for comparison. Conversion were performed using NCAT [8]. Timms Hill had a fire tower constructed on the summit in 1932, and the first ground survey of its elevation was conducted in 1934 [1] by the USGS using third order leveling. During this survey a concrete post was set on the highest point of the hill, directly underneath the fire tower, and a benchmark placed on top of the post. According to the original field notes (elevations in NGVD29), the benchmark was measured at 1953.195ft and stamped “Ogema Tower 1934 1952.868.” The field notes state that the mark projects 1.389 ft above ground at the highest point on the hill, and the ground elevation of the hill at the highest is 1951.806 ft.

It is unclear why the number 1952.868 was stamped on the marker, since this was not the measured marker height or the measured ground height.

These numbers are converted to NAVD88 for comparison purposes in Table 1.

In 1962 Pearson Hill was surveyed with third order levelling and the monument set in a concrete post [1]. The field notes state it was measured at 1950.774 ft and stamped “Pearson Hill 1962 1950.447”. It is unclear from the notes whether the highest ground or monument were measured, and the height of the concrete post is not given. It is also unclear why the monument was stamped a different number than the reported height. Representatives from USGS archives suggested it might have been a transcription error, but it’s not clear which number is correct. The 1950.774 result is assumed correct and converted to NAVD88 for comparison purposes in Table 1.

The 1979 USGS Timms Hill quad for this area lists both peaks at 1951ft based on photogrammetry conducted in 1974 and field checked in 1975 [2]. This result is the same when rounded to the nearest foot in NGVD29 or NAVD88.

In 2004 a ground survey of Timms Hill was conducted using third order leveling, and this reported an elevation of 1951.4 ft [3] (NAVD88). However, it is unclear if this is the elevation of the monument or of the ground. The NGS data sheet reports the concrete post projects 1.4ft, the same as the 1934 field notes.

In 2012 another ground survey of Timms Hill was conducted, though the method was not specified [4]. This gave a reported elevation 1952 ft (NAVD88, no decimal specified). This appears to have been rounded to the nearest foot, since it is written as “1952. ft”. The report mentions geocaching, so it was possibly surveyed using GPS. Pictures provided show the monument projecting 3ft above the ground.

No subsequesnt ground surveys of either peak have been recorded since the 2012 survey.

LiDAR

In May 2018 Lidar measurements were taken of each peak [5][6]. I downloaded the raw point cloud data and analyzed it using QGIS [9], a free open-source software package used by surveyors.

I located the highest elevation measured for each peak from the raw point cloud data. Results are given in the results section.

Measurements

On Friday evening, October 25^th I flew to Minneapolis, Minnesota, rented an SUV, and started driving into Wisconsin. I planned to survey Timms Hill first at night, since it is the more popular peak and I didn’t want anyone bothering the equipment. Then I would survey Pearson Hill in the daylight.

Timms Hill is located in a county park and the road to the trailhead is supposedly gated between sunset and sunrise. However, when I pulled up at night I found the gate open. I reached the trailhead at 3:30am and quickly hiked my gear up to the summit area a few hundred feet up the trail.

At 4am I mounted the Promark GPS unit on the benchmark with the 2.0m antenna rod. This was located directly underneath the 1932 fire tower. I then mounted the DA2 GPS on the ground just outside the fire tower so it could get a better view of the sky. The ground in between the monument and the DA2 was flat. I also used a 2.0m antenna rod for the DA2. Each site had good sky view, except for the fire towers above and some trees to the southwest. All leaves were off the tree branches, though.

With my tape measure I measured the top of the concrete post with the benchmark projected 3.0ft above the ground. This was much more than stated in the 1934 field notes and the 2004 field notes. I laid out my sleeping bag next to the tripods and took a nap until sunrise. At 8am, after four hour measurements for each GPS, I stopped logging data and packed up.

I then drove back to the main road and parked at a pullout near Bass Lake. I packed up the tripods and GPS units and hiked east to Pearson Hill, following old logging roads and bushwhacking a bit through open forest to reach the summit.

At 9am I mounted the Promark on the monument with the 2.0m antenna. I then mounted the DA2 6 ft to the west on ground level with the base of the monument as verified with my Abney level. I also used a 2.0m antenna rod. There was overhead tree cover with most leaves off the trees, but a few leaves remained. I measured the concrete post containing the Pearson benchmark projected 1.2 ft above the ground.

After 4 hours I logged the data, packed up, and hiked back down. I then drove into Michigan to survey the state highpoint candidates there.

Results

USGS

For Timms Hill I measured the monument projects 3.0ft above the highest ground. This is different than noted in the 1934 field notes, which state it projects 1.389 ft above the ground. The 1934 field notes do specify ground height, though. When converted to NAVD88 this results in a ground height of 1951.95 ft.

For Pearson Hill I measured the monument projects 1.2 ft above the highest ground. The field notes do not specify if the top of the monument or the ground were measured, and the stamped value and field notes are different, with the difference much less than 1.2ft. It’s possible there was a transcription error somewhere between the .447 and .774. If the ground was measured and the 0.774 is trusted, that would be the highest possible interpretation of the USGS measurement of the ground height. Then the final ground height in NAVD88 is 1950.93 ft. Comparing elevations, this means the third order leveling techniques by USGS measured Timms Hill 1.02 ft taller than Pearson Hill.

The 2004 and 2012 Timms Hill measurements do not specify whether they were for the monument or the ground. They are given in NAVD88 as 1951.4 ft and 1952 ft. According to USGS geographers I contacted, benchmark measurements are given for the monuments unless otherwise specified. If these were monument measurements, then this means the highest ground was measured at 1948.4 ft and 1949 ft. These were not official USGS surveys, though, so it’s possible the ground was measured.

Photogrammetry

Photogrammetry measurements from 1974 measured both peaks the same elevation to the nearest foot, 1951 ft. No error bounds were given for these measurements.

Lidar

For Timms Hill the highest datapoint was 1952.43 ft. However, this corresponds to the location of the concrete foundations on the corners of the 1982 lookout tower. These are of course not natural. The reading closest to the monument was 1951.98 ft. There are a few slightly higher readings on the north edge of the 1932 fire tower. This was an area of dense brush in 2024, so it’s possible Lidar was misclassifying brush as ground here. This is not uncommon with Lidar data.

For Pearson Hill the highest Lidar datapoint was 1950.83 ft at a location 8ft horizontal northwest of the monument. The closest point to the monument was 1950.60 ft and the closest point to the DA2 was 1950.70 ft. These measurements are all within the error bounds of Lidar measurements.

If the closest ground measurements to the monuments are used, this means Lidar measured Timms Hill 1.28 ft taller than Pearson Hill.

DA2

I processed both GPS measurements with four methods – OPUS [10], CSRS-PPP [11], TrimbleRTX [12], and Trimble Business Center.

OPUS is a free software tool provided by the US government for processing dGPS data. CSRS-PPP is a service provided by the Canadian government that uses a slightly different processing method. TrimbleRTX is a service provided by Trimble. Trimble Business center is software available from Trimble.

I have discovered based on previous surveys that in general all four processing methods give similar accuracy when surveying on open mountain tops with unobstructed sky views. Generally I get 0.1ft vertical accuracy after a one-hour measurement for all methods, and as good as 0.05 ft accuracy for a 2hr or longer measurement.

However, OPUS generally has the highest errors when surveying under tree cover. CSRS-PPP generally gives slightly lower errors in trees. TrimbleRTX generally gives the lowest errors in tree cover.

For both Pearson Hill and Timms Hill I found for all measurements TrimbleRTX gave the smallest errors. Thus, I have given final results using this processing method.

The DA2 measured the highest ground on Timms Hill was 1951. 84 ft +/- 0.14 ft (95% confidence interval).

For Pearson Hill the overhead tree cover with leaves resulted in errors +/-0.9ft even for the standard online TrimbleRTX processing site. To get a more accurate measurement, I processed this data with Trimble Business Center and manually removed cycle slips before processing. I used the WIRL base station, the closest one to the measurement. The results were 1950.94 ft +/- 0.22 ft (95% confidence interval).

This means the DA2 measured Timms hill 0.9 ft taller than Pearson Hill.

Promark

For both peaks I mounted the Promark GPS on the monument, thus corrections need to be added to get the final elevation of the highest natural point. For Timms Hill I subtracted 3.0 ft from the measured monument height. This gave an elevation of 1951.71 ft +/- 0.16 ft (95% confidence interval). For Pearson Hill I subtracted 1.2 ft from the measurement monument height. This gave an elevation of 1950.45 ft +/- 0.1 ft (95% confidence interval). As before, the Timms Hill data was processed with the standard online TrimbleRTX site, and the Pearson Hill data was processed with Trimble Business Center to remove cycle slips. The WIRL base station was used for processing.

This means the Promark measured Timms Hill 1.26 ft taller than Pearson Hill.

Overall Results

To compare the different measurements, Figure 1 shows normalized normal distributions of each measurement using the measured mean and one-sigma standard deviation values. These plots assume normal distributions of measurements. In general, the error bounds for each measurement are reported as the two-sigma standard deviations. This figure shows that all of the Timms Hill measurements are higher than all of the Pearson measurements,. The vertical lines show the mean elevations for Timms Hill and Pearson Hill.

It is unreported what the one sigma error is for the USGS measurements, so the one-sigma error is assumed to be 0.1ft.

I additionally made box plots of the same data in Figure 2. The left curve shows Timms Hill and the right Pearson Hill. The box is centered at the mean with the edge of the box +/- one sigma and the whiskers +/- two sigma. These results again show that the Timms Hill measurements are higher than the Pearson Hill measurements. Horizontal lines show the mean measured heights for Timms and Pearson.

Finally, I plotted the elevation difference for each individual measurment method in Figure 3. This allows for fair comparison between measurments. This plot shows the USGS measurements measured Timms 1.02 ft taller than Pearson, Lidar measured Timms 1.28 ft taller than Pearson, the DA2 measured Timms 0.9 ft taller than Pearson, and the Promark measured Timms 1.26 ft taller than Pearson. The horizontal line shows the mean height difference 0f 1.1 ft Timms above Pearson.

To reach a final elevation for each peak I took the mean of the four elevations for each peak and rounded to the nearest 0.1ft. This gave the final result of Timms Hill 1951.9 ft and Pearson Hill 1950.8ft. This means the final result is Timms Hill is 1.1 ft taller than Pearson Hill and Timms Hill is the state highpoint of Wisconsin.

Discussion

The USGS, Lidar, DA2, and Promark measurements each found Timms Hill taller than Pearson Hill. These are completely independent measurements using different measurement techniques (third-order levels, Lidar, and differential GPS) all giving the same results for which peak is higher. The 2004 and 2012 measurements are consistent with the USGS, Lidar, and Promark measurements if it is assumed these were reported as ground measurements after subtracting off the monument height.

In general the Pearson Hill GPS measurements had higher error, and this is likely due to the tree cover where some leaves were still present. The Timms Hill measurements had much lower error, corresponding to minimal overhead tree cover. There were fire towers near the measurements, but these did not obstruct the sky view as much as overhad branches with leaves did on Pearson Hill.

There still exist unresolved questions about the USGS measurements, even if this does not change the fact that Timms Hill is the state highpoint.

The Timms Hill monument was 1.389 ft tall in 1934, but 3.0 ft tall in 2024.  The photograph of the monument from the 2012 NGS data sheet shows it the same height as 2024. The concrete foundation of the newer lookout tower has “1982” carved in it, which appears to signify the lookout was constructed in 1982. It’s possible during construction of the new lookout tower the ground around the monument and greater summit area was removed, or the monument was altered or replaced. This would mean only measurements since 1982 would give an accurate current height of the highest ground.

Lidar measurements east of the monument in the area of the 1982 lookout tower are of ground unnaturally moved to create a foundation for the tower. Thus these measurements should not count towards the elevation of the highest ground. Lidar can also misclassify dense brush as ground, and there was dense brush around the monument area in 2024. It’s possible this added error to Lidar measurements when Lidar data was taken.

Conclusion

Timms Hill is the state highpoint of Wisconsin at 1951.9ft NAVD88. Pearson Hill is the second highest peak in Wisconsin at 1950.8ft.

References

1. Rib Lake Benchmark Records, US Department of the Interior Geological Survey, Edition of 04/08/1983. See pdf: RibLakeVerticalControls

2. U.S. Geological Survey: Timms Hill quadrangle, Wisconsin [map]. Photogrammetry 1974. Field checked 1975. 1:250,000. United States Department of the Interior, USGS, 1979

3. “Ogema”. NGS Data Sheet, 2004. National Geodetic Survey, National Oceanic and Atmospheric Administration, United States Department of Commerce.

4. “Ogema”. NGS Data Sheet, 2004. National Geodetic Survey, National Oceanic and Atmospheric Administration, United States Department of Commerce

5. U.S. Geological Survey, 20200605, USGS Lidar Point Cloud WI_Statewide_2018_A18 1619: U.S. Geological Survey.

6. U.S. Geological Survey, 20200605, USGS Lidar Point Cloud WI_Statewide_2018_A18 1620: U.S. Geological Survey.

7. USGS, “What is Lidar and Where Can I Download It?”, https://www.usgs.gov/faqs/what-lidar-data-and-where-can-i-download-it 

8. NGS Coordinate Conversion and Transformation Tool (NCAT), https://www.ngs.noaa.gov/NCAT/

9. QGIS.org. QGIS Geographic Information System. QGIS Association. http://www.qgis.org

10. Online Positioning User Service (OPUS), NOAA, https://geodesy.noaa.gov/OPUS/

11. CSRS-PPP Precise Point Positiong processing, https://webapp.csrs-scrs.nrcan-rncan.gc.ca/geod/tools-outils/ppp.php?locale=en&_gl=1*1as3oua*_ga*NTQ2NTE3OTMwLjE3MTQ2NzgzMjU.*_ga_C2N57Y7DX5*MTcyNjA4NjU

12. Trimble Centerpoint RTX Post Processing https://trimblertx.com/UploadForm.aspx

13. Gilbertson, E., Hensley, R., Kirmse, A., Bretherton, K., Stanchak, K., “LiDAR Accuracy on North American Mountain Summits,” Progress in Physical Geography: Earth and Environment, 2025 Link: https://doi.org/10.1177/03091333251401361 (Link to free version: https://arxiv.org/abs/2511.12341 )

 

 

© 2025, egilbert@alum.mit.edu. All rights reserved.