Using GNSS Observations for Tropospheric Delay Prediction Using Artificial Intelligence
In one sentence. Neural network predicts GNSS zenith tropospheric and wet delay from GNSS and meteorological data, reaching 5 mm (ZTD) and 3.6 mm (ZWD) error for water-vapour estimates.
Abstract
GNSS technology holds significant importance across wide applications, ranging from mapping, surveying, and precise timekeeping to ship navigation. Its operational principle hinges on the accurate measurement of signal travel time, which is crucial for determining the distance between the GNSS satellite and the receiving device. However, the precision of GNSS positioning is often compromised due to various error sources that impact GNSS measurements. Among these sources, atmospheric effects are widely acknowledged as the primary contributors to spatially correlated inaccuracies in GNSS (Global Navigation Satellite System) measurements. The accuracy of zenith tropospheric delay (ZTD) and zenith wet delay (ZWD) prediction using an artificial neural network model was successfully demonstrated in this study. By combining data from GNSS observations and in-situ meteorological measurements, high-resolution water vapour data can be produced for reliable and accurate weather forecasting. The validation of the predictions revealed a mean standard deviation error of 5 mm and 3.6 mm for ZTD and ZWD, respectively. This study emphasizes the significance of estimating tropospheric wet delay in real-time weather forecasting applications.
Labels
- Topic
- Observations & remote sensing, Local & station forecasting
- Type
- Application study
- Method
- Classical ML (trees, GP, linear, kernels)
- Time scale
- 0–6 h
- Domain
- Local / station
- Data
- Station / in-situ, Other
- Status
- Research
- Identifiers
- DOI:10.21608/pserj.2023.242830.1270 OpenAlex:W4388613534
- Added
- 2026-10-22
Labels and summary are generated by Claude and reviewed weekly. Wrong? Edit this entry.