Loxcel Geomatics Logo
Sign In
Sign Up

Wednesday, 2023-May-31 — Australia 5G Site Counts

This table captures the growth of 5G sites across Australia, as reported by Australia Cellular Services:

# 5G Cellular Sites Across Australia
Telstra 1 283 1,314 4,312 4,491 4,775 4,855
Optus 1 208 808 1,264 2,024 2,886 3,188
Vodafone 0 0 86 222 1,515 2,518 2,759

Wednesday, 2023-May-3 — New Zealand Cellular Services

New Zealand Cellular Services uses a database of wireless sites published by New Zealand Radio Spectrum Management (NZRSM), a government agency. This database contains location and technical information of all cellular and fixed wireless (WISP) sites across the country. For many years, NZRSM updated this database weekly — until 2022-Dec-05.

We asked NZRSM in 2023-Jan why their database had not been updated. NZRSM replied:

We registered for their API platform. After dozens of emails and hours over the next three months we discovered that the API would not support our requirements, contrary to their advice.

New Zealand Cellular Services remains available to subscribers but remember the INFORMATION IS FROM DECEMBER 2022.

We plan to update our New Zealand database but need to find new sources of information. If you have contacts with any New Zealand wireless operators (cellular or fixed wireless) please contact us to provide an introduction.

How does this affect you, as a subscriber to New Zealand Cellular Services: The pace of change is slow for cellular networks and even slower for most fixed wireless networks. As such, the information provided by New Zealand Cellular Services should remain very relevant for the next few months and somewhat relevant to the end of 2023.

Thursday, 2023-Mar-2 — The Search for a Better Digital Elevation Model


SRTM is good, but Copernicus DEM is usually better. For terrain applications, FABDEM might be even better.


We use Digital Elevation (DEM) and Land Cover Models to identify obstructions to wireless communications over long distances. DEM quality is important to our business and the success of our customers. SRTM has been the go-to DEM for most industry applications. SRTM has served us well. Recently, competitors have appeared. Should we replace SRTM with a competitor?

But, how do we know if a competitor is better than SRTM? These are actual DEM quality claims from marketing and scientific material:

These statistics are as relevant to DEM quality as player height is to the quality of a basketball team. Height is important, but not as important as a player's many quirks & features. DEM quality involves many quirks & features that elude the statistics above. This article introduces DEM Explorer, a visualization tool that lets you explore DEM quality and discover its quirks & features.

DEM Explorer graphs the behavior of the 14 DEMs below, across 3 land covers and 4 slopes. Each curve plots the distribution of error [ elevDEM - elevLiDAR ] between a DEM and highly accurate LiDAR ground truths. Pan, zoom, pinch & swipe the graph to discover more quirks & features.

DEM Contenders

DEM Version Released Notes
ASTER v003 2019-06 Put it out to pasture.
AW3D30 v3.2 (Feb 2022) 2022-02
COPernicus DEM 30 DGED 2022_1 2023-01 Pixels 32 bit float
COPernicus DEM 30 0.5m DGED 2022_1 2023-01 Pixels rounded to 0.5m
COPernicus DEM 30 1m DGED 2022_1 2023-01 Pixels rounded to 1m
COPernicus DEM 90 DGED 2022_1 2023-01 Pixels 32 bit float
COPernicus DEM 90 1m DGED 2022_1 2023-01 Pixels rounded to 1m
FABDEM V1-2 2023-01 COP30 with reduced forest & building bias
ICESat-2_2 v005 2023-02 h_te_uncertainty < 2
ICESat-2_10 v005 2023-02 h_te_uncertainty < 10
MERIT v1.0.3 2018-10 SRTM, with less forest bias
NASADEM HGT v001 2020 Reprocessed SRTM
SRTM SRTMGL1 v003 2016 Very popular
TDX90 v3 2016 Foundation of COP30 and COP90

Try for Yourself

Click any graph on the right to launch DEM Explorer and see


We compare each DEM listed above to billions of LiDAR ground truths with a point density > 6 / m2. Each colored curve on the graph is a distribution of error, created by comparing one DEM to all LiDAR ground truths with the same land cover and slope; eg. forest with moderate slope. Each curve stresses the DEM in a different way, teasing out biases. A smooth curve requires at least one million LiDAR ground truths; most curves use many more (billions in some cases) producing the smooth curves you see on the right.

These LiDAR ground truths have a vertical accuracy better than the height of a chipmunk (5 to 10 cm). We quote RMSE, MAE and other statistics to 0.1m precision; a higher precision captures only terrain noise, like chipmunks, acorns and other ephemeral clutter.

ESA WorldCover 10m 2021 V200 identifies a land cover for each LiDAR elevation: Grass / Crop, Forest or Developed. (We combine Grassland and Cropland, as they present remote sensing with a similar, short and easily permeable surface.)

Slope is calculated from a high-resolution 0.5m elevation grid created from the LiDAR ground truths, providing the most accurate ground slope possible:

LabelSlope (%)
Level< 1
Gentle1 to 4
Moderate4 to 12
Steep12 to 100

Elevation Normalization (Geoid to Ellipsoid)

LiDAR and DEM surveys capture ellipsoidal elevation which are later converted to geoidal (eg. EGM96, EGM2008, NAVD88, CGG2013) for public use. Our analysis require all LiDAR and DEM elevations normalized to the WGS84 ellipsoid. Normalization uses a geoid grid and an interpolation method; each grid size & interpolation combination produces different results. Normalization error occurs if the combination we use differs from what was used when the elevation data was packaged for public use. This error can vary from centimeters to meters.

A geoid's continuous surface is defined by spherical harmonic coefficients. These coefficients are too computationally expensive to work with directly, so they are digitized once into a grid of pixels, which approximate the geoid's surface, and interpolated, on demand, to obtain geoid offsets.

Our normalization (from geoid to WGS84) should use the same grid size and interpolation used when the DEM was created. However, only NASADEM publishes these details (ie. 15 arcsecond grid with linear interpolation). We applied various geoid grid sizes and interpolation methods to COP90 to discover how it was derived from TanDEM-X 90m (ie. 60 arcsecond grid with linear interpolation); we assume the same for COP30 but cannot confirm because TanDEM-X does not publish a 1 arcsecond spacing DEM. We used a 60 arcsecond grid and spline interpolation for other DEMs, as that is the interpolation method used by the US National Geospatial Intelligence Agency (masters of the geoid) in their calculations.

These details are important wherever the geoid undulates strongly, such as Hawaii, a place we are currently studying.


DEM quality is not a constant and depends on use-case, nature-of-bias, budget, license terms, file size and coverage area.

DEM Explorer can help you understand a DEM's nature-of-bias from comparisions with 627 billion LiDAR ground truths in Southern Ontario and 12.3 billion more in Newfoundland, Canada.

Marketing and scientific literature often use RMSE (root mean square error) as a proxy for DEM quality. RMSE must be used with caution, because a few bad apples can spoil the results. To that end, DEM Explorer provides >10m and >20m threshold statistics, measuring the percentage of bad apples (ie. percentage of error above 10m and 20m) which sends RMSE soaring. DEM Explorer also provides MAE (mean absolute error) which is less sensitive to extreme outliers. But, RMSE or MAE — alone — are as much a sign of DEM quality as player height is to basketball team quality.

What's the answer to replacing SRTM? Switching DEMs is not a simple exercise. A new DEM brings its own quirks & features that will improve some things and worsen others. Will the mix of quirks & features net a positive outcome?

Copernicus DEM provides much more accurate surface elevations, useful for our work in wireless propagation analysis. FABDEM is a derivative of Copernicus DEM that reduces this surface bias, which you need for flood analysis. FABDEM performs this task well, at a cost of some negative bias. As well, FABDEM has restrictive terms of license which a commercial application must consider.

We use DEMs for wireless propagation analysis, which favors a DEM that captures all surface clutter (forests, shrubs, but not chipmunks or acorns). Other use-cases, like floodplain analysis, need a no-clutter DEM. These and other quirks & features are what DEM Explore can help you discover, on your search for a better DEM.

These graphs compare DEMs to LiDAR ground elevations.
Curves depict distribution of error
[ elevDEM - elevLiDAR ]
> Click any graph to explore further <

Figure 1: One RMSE value cannot capture DEM quality

For Southern Ontario, Canada, distribution graph of error of SRTM DEM in [ forest moderate slope, rmse 5.4 mae 4.2 ] [ developed moderate rmse 3.0, mae 2.1 ] [ grass/crop level rmse 1.7, mae 1.3 ]
ColorDEMLand coverSlopeRMSE

This range of RMSE values — for one DEM — shows the folly of representing DEM quality with a single RMSE value. Yet scientific and marketing literature does this.

Figure 2: Copernicus DEM quality eclipses NASADEM in low-clutter terrain

For Southern Ontario, Canada, distribution graph of error of Copernicus DEM 30 vs NASADEM grass/crop level slope [ COP30 rmse 0.9, mae 0.4 ] [ NASADEM rmse 1.7, mae 1.3 ]

NASADEM's shallow curve isn't a sign of how bad it is (NASADEM is a good DEM). Instead, it's a sign of how good Copernicus DEM is in certain situations.

Figure 3: Copernicus DEM can't see the forest floor

For Southern Ontario, Canada, distribution graph of error of Copernicus DEM 30 vs FABDEM forest moderate slope [ COP30 rmse 8.8, mae 7.3 ] [ FABDEM rmse 3.8, mae 2.8 ]

COP30's yellow tail to the right of the y-axis is forest clutter; a benefit to radio propagation analysis but a detriment to floodplain analysis. FABDEM's green curve reduces clutter bias, at a cost of bias left of the y-axis.

Figure 4: Copernicus DEM does better in Newfoundland, Canada.

For Avalon Peninsula, Newfoundland, Canada, distribution graph of error of Copernicus DEM 30 vs FABDEM forest moderate slope [ COP30 rmse 8.8, mae 7.3 ] [ FABDEM rmse 3.8, mae 2.8 ]

Figures 3 & 4 show different RMSE & MAE values for the same DEM in forests with moderate slope. The only difference is place, expanding on observations in Figure 1.

Figure 5: Effects of rounding DEM pixels (no effect)

For Southern Ontario, Canada, distribution graph of error of Copernicus DEM 30 with float and integer pixels in level forest surfaces [ float and integer rmse 7.3, mae 5.7 ]

Rounding Copernicus DEM pixels from float to integer significantly improves compression ratios, benefiting resource constrained devices. Quality is not compromised when rounding pixels in forests with level terrain.

Figure 6: Effects of rounding DEM pixels (some effect)

For Southern Ontario, Canada, distribution graph of error of Copernicus DEM 30 with float and integer pixels in level grass/crop terrain [ float rmse 0.9, mae 0.4 ] [ integer rmse 1.0 mae 0.5 ]

The effects of rounding are slightly worse in Grass / Crop surfaces with level terrain.

Figure 7: ICESat-2, as a sparse-DEM, in forested, moderate slope areas

For Southern Ontario, Canada, distribution graph of error of ICESat-2 with h_te_uncertainty less than 2 and 10 [ lt 2 rmse 1.17, mae 0.65 ] [ lt 10 rmse 1.73, mae 0.76 ]
Red< 21.170.65-0.021.16
Green< 101.730.76-0.141.73

ICESat-2 ATL08 segments are used in other studies as ground truths to assess DEM quality. We turned the tables, creating two sparse-DEMs from 68,600 km2 of ATL08 segments and compared them to our higher accuracy LiDAR. DEM Explorer shows this sparse-DEM to be more far accurate than any global DEM.

Figure 8: COP90 (green) and TanDEM-X (yellow) share DNA

For Avalon Peninsula, Newfoundland, Canada, distribution graph of error of Copernicus DEM 90 and TanDEM-X 90 for level developed terrain [ COP90 rmse 1.2, mae 0.9 ] [ TDX90 rmse 1.6, mae 0.9 ]

COP30 and COP90 are derived from TanDEM-X. COP90 and TDX90 are often identical. This graph shows divergence for negative bias, possibly due to manual editing of COP90.

Friday, 2023-Jan-20 — January Update (Canada)

The graphs below plot the most recent 18 months of ISED SMS snapshots, by channel count (top), occupied spectrum (middle) and site count (bottom), for the three national (left) and four regional (right) carriers.

Rogers' graphs (red) naturally trend upwards, capturing the growth of its network. Telus' (green) and Bell's (blue) graphs see-saw up and down, hilighting the inconsistencies of these snapshots.

We fix these snapshots to ensure Canada Cellular Services provides the most accurate account of Canada's wireless networks available.

 Site Counts
Filter Dec-22Jan-23Increase
700 MHz B12 5,18314,2789,095
Telus/Bell LTE 22,86424,9382,074
Bell 8,33710,1011,764
AWS-3 6,2097,7031,494
Telus/Bell 5G 6981,615917
700 MHz B29 4,8135,012199
Rogers 5G 4,5114,58170

Channel Count
Rogers, Telus, Bell channel count graph for past 18 months
Channel Count
Freedom, Videotron, SaskTel, Eastlink channel count graph for past 18 months

Occupied Spectrum (GHz)
Rogers, Telus, Bell occupied spectrum graph for past 18 months
Occupied Spectrum (GHz)
Freedom, Videotron, SaskTel, Eastlink occupied spectrum graph for past 18 months

Site Count
Rogers, Telus, Bell site count graph for past 18 months
Site Count
Freedom, Videotron, SaskTel, Eastlink site count graph for past 18 months

Current     2022     2021     2020     2019     2018     2017     2016     2015     2014     2013     2012