Marathon Running with the “Apple Watch Ultra 3”

By /
Marathon Running with the Apple Watch Ultra 3

Apple Watch Ultra 3: GPS Accuracy for Marathon Runners in the City

Running a marathon in a big city – such as New York, Chicago, Tokyo, or even a densely built European city center – is not only a huge physical and psychological challenge for the athlete but also an extremely complex technological challenge for their wearable devices. One of the biggest problems faced by long-distance runners is accurately maintaining their running pace. While traditional smartwatches perform well in open areas, once they enter streets surrounded by skyscrapers, navigational chaos begins: the pace on the screen suddenly drops, the displayed distance becomes distorted, and after finishing, the running trajectory shown in the app resembles zigzags crossing building walls.

The new Apple Watch Ultra 3, equipped with an updated GPS architecture and a powerful S11 processor, is designed specifically for such extreme urban conditions. In this article, we will thoroughly examine the main navigation problems in the city and the technical solutions that the Ultra 3 employs to maintain impeccable accuracy.

1. Navigation Enemy No. 1: Urban Canyon and Multipath Effect

To understand why GPS accuracy disappears when running in the city, we need to look at the laws of physics. Global Navigation Satellite Systems (GNSS), such as GPS, GLONASS, Galileo, QZSS, and BeiDou, transmit radio signals from approximately 20,000 km above the Earth. When these signals reach the ground, they are extremely weak (their power is often compared to that of a 25-watt light bulb seen from another continent).

When a runner finds themselves on a street flanked by tall buildings (this phenomenon is known in navigation science as Urban Canyon), the direct visibility between the watch and the satellite – Line-of-Sight (LOS) – is often blocked. Instead, the watch’s antenna picks up signals that bounce off the glass and metal facades of skyscrapers.

This phenomenon is called the Multipath effect, and the reflected signals are referred to as Non-Line-of-Sight (NLOS). Since the reflected signal travels a longer distance than the direct one, the receiver mistakenly calculates that the satellite is farther away than it actually is. As a result, the runner’s position on the map is “pushed” several dozen meters to the side, and the watch incorrectly registers a sudden increase or decrease in speed.

2. Precision Dual-Frequency Architecture: L1 and L5 Standards

Older generation watches relied on only one radio frequency – L1. While the L1 frequency is reliable in open areas, it is highly sensitive to obstacles and reflections. The Apple Watch Ultra 3 addresses this issue by using an integrated Precision dual-frequency GPS (L1 + L5) system.

The L5 frequency is an important weapon against the Multipath effect. This frequency was initially developed for aviation purposes. Its main advantages include:

  • Higher Chipping Rate: The modulation of the L5 signal is significantly faster. This means that the signal “pulses” are shorter, making it much easier for the receiver’s correlators to distinguish the true, original signal from its reflection, which arrives just a few nanoseconds later.
  • Better Penetration and Throughput: The L5 signal, due to its wave properties and higher transmission power, more easily overcomes tree canopies and signals reflected from building structures.

The Ultra 3 receiver continuously compares data received on L1 and L5 frequencies. If the L1 signal shows anomalies (due to reflection), the algorithm prioritizes L5 data, thus filtering out spatial noise.

3. Sensor Fusion: When Satellites Stop Working

Even with the L5 frequency, there can be moments in a dense Urban Canyon when all direct signals are blocked. This is where the Ultra 3 employs Sensor Fusion methodology – a process where the chip combines GNSS data with inertial sensors (IMU).

The watch integrates a highly sensitive accelerometer and gyroscope. As the runner moves, these MEMS (micro-electromechanical systems) sensors measure the frequency of arm swings, stride length, and body movement vector.

When the Multipath effect becomes too strong and the GPS position error (Error RMSE) exceeds allowable limits, the Kalman filter in the Ultra 3 system automatically reduces trust in satellite data and begins to rely on Dead Reckoning algorithms (calculating movement based on inertia). In other words, the watch maintains your pace and distance solely based on your movement biomechanics until it “catches” a clean Line-of-Sight signal again.

4. Artificial Intelligence and 3D Modeling: The Power of the S11 Processor

Hardware is only half the job. The greatest advantage of the Ultra 3 lies in the new S11 chip set and its Neural Engine cores. In simple receivers, Multipath mitigation is performed based on basic mathematical models. Apple has implemented a radical solution – location determination based on machine learning methods.

  • Real-Time Signal-to-Noise Ratio (SNR) Analysis: The S11 processor analyzes the quality of each received signal in microseconds and discards satellites whose data is recognized as reflections.
  • 3D Map Integration (3D Digital Surface Model): Using basic Apple Maps topography data, algorithms “know” in advance that the runner is entering a street with tall buildings. The system proactively adjusts Stochastic model settings, anticipating greater signal reflections, and compensates for errors accordingly. This level of spatial awareness allows the watch to ignore impossible trajectory changes (e.g., a runner cannot suddenly be behind a building 50 meters away in one second).

5. What Does This Mean for the Runner on the Course?

Theoretical innovations materialize into very practical benefits during a marathon:

  • Accurate and Consistent Pace (Current Pace): Unlike older watches that would show a pace fluctuating from 4:15 min/km to 6:30 min/km in the city center, the Ultra 3 screen provides a stable pace that reflects your actual efforts. This is vital for planning energy resources.
  • Consistent Kilometer Splits: There are no “phantom” kilometers that count due to GPS fluctuations back and forth.
  • Better Battery Management: Although dual-frequency tracking requires more energy, the efficiency of the S11 processor allows it to process massive Sensor Fusion data streams with minimal energy consumption.

Conclusion

The Apple Watch Ultra 3 rewrites the standards for wearable technology for marathon runners. By addressing the challenges posed by Urban Canyon and the Multipath effect through L1+L5 frequency architecture, Sensor Fusion, and intelligent S11 processing, this device allows the athlete to focus on what matters most – the run itself. No more guessing your true speed or being disappointed by inaccurate statistics after a tough workout in the city.

Sources and Recommended Literature

  • u-blox (2026). Real-Time Multipath Mitigation with Sensor-Aided Long Coherent Integration (SALI). Access via navigation engineering databases.
  • Liu, M., et al. (2023). Multi-Phase Fusion for Pedestrian Localization Using Mass-Market GNSS and MEMS Sensors. PMC. Available online.
  • Zhang, Y., & Chen, L. (2025). High-Precision Multi-Source Fusion Navigation Solutions for Complex and Dynamic Urban Environments. MDPI Remote Sensing (Available via Google Scholar).
  • Hsu, L. T., et al. (2015/2024). GNSS positioning with multipath simulation using 3D surface model in Urban Canyon. Access via ResearchGate / Google Scholar.
  • Apple Inc. (2025/2026). Apple Watch Ultra 3 – Technical Specifications (L1 and L5 Dual-Frequency GPS). Official documentation.