River surface velocity estimation using UAS-borne Doppler radar and continuous wavelet transform

Accurate measurement of river surface velocity is essential for hydrological monitoring, flood forecasting, and water resource management. In contrast with traditional in-situ point measurements using electromagnetic current meters, remote sensing techniques offer significant advantages for river surface velocity estimation, including rapid data acquisition, lower operational costs, and contactless operation. Based on Unmanned Aerial Systems (UAS) equipped with Doppler radar becomes more attractive due to it is suitable for real-time velocity determination and has fewer limitations.

The UAS-borne Doppler radar estimates river surface velocity by detecting the frequency shift of backscattered microwaves, with the drone operating at a controlled hover altitude of approximately 4 meters above the water surface during field measurements. However, the propeller wash (propwash) generated by the UAS distorts the radar return signal, resulting in a composite velocity measurement that combines the true river surface flow with the locally induced airflow velocity. To isolate the true river velocity, we introduce a bidirectional observation scheme in which the same surface footprint is measured from two opposing directions. The Continuous Wavelet Transform (CWT) algorithm is employed to extract the mixed velocity components from each directional dataset. By analytically reconciling these bidirectional measurements, the downwash contribution is effectively removed, thereby yielding a refined estimate of the true river surface velocity.

In this study, river surface velocity was analysed across more than 50 cross-sections spanning four distinct rivers, covering a broad velocity range from 20 cm/s to 250 cm/s. To validate the velocity estimates obtained from the UAS-Doppler radar and CWT method, comparisons were made against in-situ measurements collected using instruments such as the OTT MF Pro, Acoustic Doppler Current Profilers (ADCP), and flow trackers. Quantitative analysis confirmed that the UAS-Doppler radar system provides reliable river flow velocity measurements while offering enhanced efficiency in post-processing workflows.